CN107922495B

CN107922495B - Bispecific antibody constructs that bind to EGFRVIII and CD3

Info

Publication number: CN107922495B
Application number: CN201680044931.9A
Authority: CN
Inventors: T·劳姆; P·库弗; D·劳; M·明兹; I·赫尔曼; P·霍夫曼
Original assignee: Amgen Research Munich GmbH
Current assignee: Amgen Research Munich GmbH
Priority date: 2015-07-31
Filing date: 2016-08-01
Publication date: 2022-05-10
Anticipated expiration: 2036-08-01
Also published as: CR20180065A; PL3328892T3; LT3328892T; IL256872A; CY1124384T1; EP3912999A1; CO2018000887A2; AU2016302683B2; CA2991278C; CA2991278A1; WO2017021370A1; TN2017000548A1; AU2022291524A1; SG10202109625YA; JO3788B1; PH12018500156A1; PE20181013A1; CN107922495A; US20170029512A1; DK3328892T3

Abstract

The present invention relates to a bispecific antibody construct comprising a first binding domain that binds to human EGFRVIII on the surface of a target cell and a second binding domain that binds to human CD3 on the surface of a T cell. Furthermore, the invention provides polynucleotides encoding the antibody constructs, vectors comprising the polynucleotides and host cells transformed or transfected with the polynucleotides or vectors. In addition, the invention provides a method for producing the antibody construct of the invention, a medical use of the antibody construct and a kit comprising the antibody construct.

Description

Bispecific antibody constructs that bind EGFRVIII and CD3

EGFRvIII is one of several EGFR variants caused by gene rearrangement accompanied by EGFR gene amplification. EGFRvIII consists of an in-frame deletion of 267 amino acids in the extracellular domain of EGFR. EGFRvIII is the most common variant of the Epidermal Growth Factor (EGF) receptor in human cancers. During gene amplification, the extracellular domain was deleted by 267 amino acids, forming a tumor-specific neo-epitope neo-junction that could be used as a monoclonal antibody. This variant of the EGF receptor promotes tumor progression by constitutive signaling in a ligand independent manner. Expression of EGFRvIII on any normal tissue has not been found. The excellent tumor selectivity of EGFRvIII expression makes EGFRvIII an ideal antigen for targeting with BiTE antibody constructs. The contribution of EGFRvIII to tumor progression suggests a dependence of cancer cells on EGFRvIII expression and thus further supports its suitability as a target.

EGFRvIII has been shown to be frequently expressed in two types of malignant central nervous system tumors (i.e., glioblastoma multiforme and anaplastic astrocytoma). There is an unmet significant medical need for both diseases. This can be exemplified by poor overall survival under standard of care, with a 2-year overall survival of 13.6% for glioblastoma multiforme and a 5-year overall survival of 25.9% for anaplastic astrocytoma. Currently, the standard of care for glioblastoma multiforme consists of surgical resection of tumors, which is most often limited by the diffuse growth pattern of the tumor and the need to preserve functionally important areas of the brain. The surgery is followed by adjuvant irradiation and chemotherapy. For glioblastoma multiforme and anaplastic astrocytoma treated according to current standard of care, recurrence is normal and ultimately fatal outcome in almost all patients.

Expression of EGFRvIII in glioblastoma multiforme and anaplastic astrocytomas constitutively activates the PI3 kinase signaling pathway and correlates with prognosis of exacerbations in glioblastoma multiforme and anaplastic astrocytomas. EGFRvIII has also been described to co-express with CD133 and define Cancer stem cell populations in glioblastoma multiforme (Emlet et al, Cancer Res, 2014, 74 (4): 1238-49). Furthermore, it has been demonstrated that EGFRvIII can be transferred to the cell surface of antigen-negative tumor cells by the mechanism of intercellular antigen transfer. By this mechanism, the expression heterogeneity of EGFRvIII, which has been shown in some glioblastoma multiforme, can be overcome as a barrier to therapeutic efficacy, particularly with EGFRvIII-specific BiTE antibody constructs, providing a highly efficient cytotoxic profile that compensates for the effects of potentially low-level target antigens resulting from intercellular antigen transfer.

EGFRvIII expression is also described in several other tumor entities (Wikstrand, CJ. et al, Cancer Research 55 (14): 3140-3148 (1995); Ge H. et al, Int J cancer.98 (3): 357-61 (2002); Moscatello, G. et al, Cancer Res.55 (23): 5536-9 (1995); Garcia de Palazzo, IE. et al, Cancer Res.53 (14): 3217-20 (1993); olapad-olaopapa, EO. et al, Br J cancer.82 (1): 186-94 (2000)). Thus, given the precise tumor selectivity of EGFRvIII, treatment with EGFRvIII-specific BiTE antibody constructs may also be beneficial for other selected cancer types or subtypes. The potential selection of the type of cancer, subtype or patient to be treated can be guided by various methods of testing the expression of EGFRvIII in tumors.

As there remains a need for more options that can be used to treat solid tumor diseases associated with EGFRVIII overexpression (e.g., glioblastoma, astrocytoma, medulloblastoma, breast cancer, non-small cell lung cancer, ovarian cancer, prostate cancer, central nervous system cancer), means and methods are provided to address this problem by bispecific antibody constructs having a binding domain for EGFRVIII on the surface of a tumor target cell and a second binding domain for CD3 on the surface of a T cell.

Accordingly, in a first aspect, the present invention provides a bispecific antibody construct comprising a first binding domain that binds to human and cynomolgus EGFRVIII on the surface of a target cell and a second binding domain that binds to human CD3 on the surface of a T cell, wherein said first binding domain comprises the amino acid sequence as set forth in SEQ ID NO:157 and the polypeptide as set forth in SEQ ID NO: 158.

It must be noted that, as used herein, the singular forms "a", "an" and "the" include the plural forms unless the context clearly dictates otherwise. Thus, for example, reference to "an agent" includes one or more of such different agents, and reference to "the method" includes reference to equivalent steps and methods known to those of ordinary skill in the art that may be modified or substituted for the methods described herein.

The term "at least" used in connection with a series of elements is to be understood as referring to each element in the series unless otherwise indicated. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the present invention.

Wherever used herein, the term "and/or" includes "and", "or" and "all or any other combination of elements connected by the term" are meant.

As used herein, the term "about" or "approximately" means within ± 20%, preferably within ± 15%, more preferably within ± 10%, and most preferably within ± 5% of a given value or range.

Throughout this specification and the claims which follow, unless the context requires otherwise, the word "comprise", and variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps. The term "comprising" may be substituted with the term "comprising" or "including" as used herein, or sometimes with the term "having" as used herein.

As used herein, "consisting of" does not include any elements, steps, or components not specified in the claimed elements. As used herein, "consisting essentially of does not exclude materials or steps that do not materially affect the basic and novel characteristics of the requirements.

In each case herein, any of the terms "comprising," consisting essentially of, "and" consisting of may be replaced with either of the other two terms.

The term "antibody construct" refers to a molecule in which the structure and/or function is based on that of an antibody, e.g., a full-length or intact immunoglobulin molecule. Thus, the antibody construct is capable of binding to its specific target or antigen. Furthermore, the antibody construct according to the invention comprises the minimum structural requirements of the antibody allowing target binding. This minimum requirement may, for example, be defined as the presence of at least three light chain CDRs (i.e. CDR1, CDR2 and CDR3 of the VL region) and/or three heavy chain CDRs (i.e. CDR1, CDR2 and CDR3 of the VH region), preferably all 6 CDRs. The antibodies on which the constructs of the invention are based include, for example, monoclonal, recombinant, chimeric, deimmunized, humanized and human antibodies.

Full-length or intact antibodies are encompassed within the definition of "antibody constructs" according to the invention, which also includes camelid antibodies and other immunoglobulin antibodies produced by biotechnological or protein engineering methods or treatments. These full-length antibodies can be, for example, monoclonal antibodies, recombinant antibodies, chimeric antibodies, deimmunized antibodies, humanized antibodies, and human antibodies. Also within the definition of "antibody construct" are fragments of full-length antibodies, such as VH, VHH, VL,(s) dAb, Fv, Fd, Fab ', F (ab') 2 or "r IgG" ("half-antibodies"). The antibody construct according to the invention may also be a modified fragment of an antibody (also referred to as antibody variant), such as an scFv, di-scFv or bi(s) -scFv, scFv-Fc, scFv-zipper, scFab, Fab2, Fab3, diabody, single-chain diabody, tandem diabody (Tandab's), tandem di-scFv, tandem tri-scFv; "minibody", the structure of which is exemplified below: (VH-VL-CH3)2, (scFv-CH3)2, ((scFv)2-CH3+ CH3), ((scFv)2-CH3) or (scFv-CH3-scFv) 2; multiple antibodies such as triabodies or tetrabodies; and single domain antibodies, such as nanobodies or single variable domain antibodies comprising only one variable domain (which may be VHH, VH or VL) that specifically binds an antigen or epitope independently of the other V regions or domains.

The binding domain may typically comprise an antibody light chain variable region (VL) and an antibody heavy chain variable region (VH); however, it need not contain both. For example, the Fd fragment has two VH regions, and often retains some of the antigen binding function of the entire antigen binding domain. Other examples of forms of antibody fragments, antibody variants or binding domains include: (1) fab fragments, monovalent fragments with VL, VH, CL and CH1 domains; (2) a F (ab') 2 fragment, a bivalent fragment having two Fab fragments linked by a disulfide bond in the hinge region; (3) an Fd fragment having two domains of VH and CH 1; (4) (ii) an Fv fragment having VL and VH domains of a single arm of an antibody; (5) dAb fragments having a VH domain (Ward et al, (1989) Nature 341: 544-546); (6) an isolated Complementarity Determining Region (CDR); and (7) single chain fv (scFV), the latter being preferred (e.g., derived from a scFV library). Examples of embodiments of antibody constructs according to the invention are described in, for example, WO 00/006605, WO 2005/040220, WO 2008/119567, WO 2010/037838, WO 2013/026837, WO 2013/026833, US 2014/0308285, US 2014/0302037, WO 2014/144722, WO 2014/151910 and WO 2015/048272.

Furthermore, the definition of the term "antibody construct" includes monovalent, bivalent and multivalent constructs, and thus includes monospecific constructs that specifically bind to only one antigenic structure, as well as bispecific and multispecific constructs that specifically bind to more than one antigenic structure (e.g., two, three or more) through different binding domains. Furthermore, the definition of the term "antibody construct" includes molecules consisting of only one polypeptide chain, as well as molecules consisting of more than one polypeptide chain, which chains may be identical (homodimers, homotrimers or homooligomers) or different (heterodimers, heterotrimers or heterooligomers). Examples of such Antibodies and variants or derivatives thereof are described, inter alia, in Harlow and Lane, Antibodies a laboratory manual, CSHL Press (1988) and Using Antibodies: a laboratory manual, CSHL Press (1999); kontermann and Dubel, Antibody Engineering, Springer, 2 nd edition, 2010 and Little, Recombinant Antibodies for immunology 2009.

The antibody construct of the invention is preferably an "in vitro generated antibody construct". The term refers to an antibody construct according to the above definition in which all or part of the variable region (e.g., at least one CDR) is produced in a non-immune cell selection (e.g., in vitro phage display), protein chip, or any other method in which the antigen binding ability of a candidate sequence can be tested. Thus, the term preferably does not include sequences that result solely from genomic rearrangements in animal immune cells. A "recombinant antibody" is an antibody produced by using recombinant DNA techniques or genetic engineering.

As used herein, the term "monoclonal antibody (mAb)" or monoclonal antibody construct refers to an antibody obtained from a population of substantially homogeneous antibodies (i.e., the individual antibodies comprising the population are identical except for possible naturally occurring mutations and/or post-translational modifications (e.g., isomerization, amidation) that may be present in minor amounts). Monoclonal antibodies are highly specific against a single antigenic site or determinant on an antigen, as compared to conventional (polyclonal) antibody preparations which typically comprise different antibodies directed against different determinants (or epitopes). In addition to their specificity, monoclonal antibodies have the advantage that they are synthesized by hybridoma culture and are therefore uncontaminated by other immunoglobulins. The modifier "monoclonal" indicates the character of the antibody as obtained from a substantially homogeneous population of antibodies, and is not to be construed as requiring production of the antibody by any particular method.

For the preparation of monoclonal antibodies, any technique that provides antibodies produced by continuous cell line culture can be used. For example, monoclonal antibodies to be used may be identified by a method described by Koehler et al, Nature, 256: 495(1975) or may be prepared by recombinant DNA methods (see, e.g., U.S. patent No. 4,816,567). Examples of other techniques for generating human Monoclonal Antibodies include the trioma technique, the human B-cell hybridoma technique (Kozbor, Immunology Today 4(1983), 72) and the EBV-hybridoma technique (Cole et al, Monoclonal Antibodies and Cancer Therapy, Alan R.Liss, Inc. (1985), 77-96).

Standard methods (e.g., enzyme-linked immunosorbent assay (ELISA) and surface plasmon resonance (BIACORE)) can then be used^TM) Assay) screening of hybridomas for the identification of production of specific antigensOne or more hybridomas that specifically bind the antibody. Any form of the relevant antigen may be used as the immunogen, e.g., recombinant antigen, naturally occurring form, any variant or fragment thereof, and antigenic peptides thereof. Surface plasmon resonance used in BIAcore systems can be used to increase the efficiency of phage Antibodies that bind to epitopes of target antigens (e.g., EGFRVIII or CD3 epsilon) (Schier, Human Antibodies hybrids 7(1996), 97-105; Malmborg, J.Immunol. methods 183(1995), 7-13).

Another exemplary method of making monoclonal antibodies includes screening protein expression libraries, such as phage display libraries or ribosome display libraries. Phage display is described, for example, in Ladner et al, U.S. patent nos. 5,223,409; smith (1985) Science 228: 1315-1317; clackson et al, Nature, 352: 624-: 581-597 (1991).

In addition to the use of display libraries, non-human animals, e.g., rodents (e.g., mice, hamsters, rabbits, or rats) can be immunized with the relevant antigen. In one embodiment, the non-human animal comprises at least a portion of a human immunoglobulin gene. For example, large segments of human Ig (immunoglobulin) loci can be used to engineer mouse strains that lack mouse antibody production. Using hybridoma technology, antigen-specific monoclonal antibodies derived from genes with the desired specificity can be generated and selected. See, e.g., XENOMOUSE^TM(ii) a Green et al (1994) Nature Genetics 7: 13-21; US 2003-0070185; WO 96/34096 and WO 96/33735.

Monoclonal antibodies can also be obtained from non-human animals and then modified, e.g., humanized, de-immunized, chimeric, etc., using recombinant DNA techniques known in the art. Examples of modified antibody constructs include humanized variants of non-human antibodies, "affinity matured" antibodies (see, e.g., Hawkins et al, J.mol.biol.254, 889-896(1992) and Lowman et al, Biochemistry 30, 10832-10837(1991)) and antibody mutants with altered effector functions (see, e.g., U.S. Pat. No. 5,648,260; Kontermann and Dubel (2010), supra; and Little (2009), supra).

In immunology, affinity maturation is the process in which B cells produce antibodies with increased affinity for an antigen during an immune response. With repeated exposure to the same antigen, the host will produce antibodies of successively higher affinity. As with the natural prototype, in vitro affinity maturation is based on the principles of mutation and selection. In vitro affinity maturation has been successfully used to optimize antibodies, antibody constructs, and antibody fragments. Random mutations within the CDRs are introduced using radiation, chemical mutagens, or error-prone PCR. In addition, genetic diversity can be increased by chain shuffling. Two or three rounds of mutation and selection using display methods such as phage display typically produce antibody fragments with affinities in the low nanomolar concentration range.

Preferred types of amino acid substitution variants of the antibody construct involve substituting one or more hypervariable region residues of a parent antibody (e.g. a humanized or human antibody). Typically, the resulting variant selected for further development will have improved biological properties relative to the parent antibody from which it was produced. A convenient method for generating such surrogate variants involves affinity maturation using phage display. Briefly, several hypervariable region sites (e.g., 6-7 sites) were mutated to generate all possible amino acid substitutions at each site. The antibody variants thus generated are displayed in a monovalent fashion from filamentous phage particles as fusions of the gene III product of M13 encapsulated within each particle. The phage display variants are then screened for their biological activity (e.g., binding affinity), as disclosed herein. To identify candidate hypervariable region sites to be modified, alanine scanning mutagenesis can be performed to identify hypervariable region residues that significantly contribute to antigen binding. Alternatively or additionally, it may be beneficial to analyze the crystal structure of the antigen-antibody complex to identify contact points between the binding domain and, for example, human EGFRVIII. Such contact residues and adjacent residues are candidates for substitution according to the techniques set forth herein. Once such variants are generated, the set of variants is screened as described herein, and antibodies with superior properties can be selected in one or more relevant assays for further development.

The monoclonal antibodies and antibody constructs of the invention specifically include "chimeric" antibodies (immunoglobulins) in which a portion of the heavy and/or light chain is identical to or homologous to corresponding sequences in antibodies derived from a particular species or belonging to a particular antibody class or subclass, while the remainder of the chain is identical to or homologous to corresponding sequences in antibodies derived from another species or belonging to another antibody class or subclass, as well as fragments of such antibodies, so long as they exhibit the desired biological activity (U.S. Pat. No. 4,816,567; Morrison et al, Proc. Natl.Acad.Sci.USA, 81: 6851-6855 (1984)). Chimeric antibodies of interest herein include "primatized" antibodies comprising variable domain antigen binding sequences derived from a non-human primate (e.g., old world monkey, ape, etc.) and human constant region sequences. Various methods for making chimeric antibodies have been described. See, e.g., Morrison et al, proc.natl.acad.scl u.s.a.81: 6851, 1985; takeda et al, Nature 314: 452, 1985; cabilly et al, U.S. Pat. No. 4,816,567; boss et al, U.S. patent No. 4,816,397; tanaguchi et al, EP 0171496; EP 0173494; and GB 2177096.

Antibodies, antibody constructs, antibody fragments or antibody variants can also be modified by specific deletion of human T cell epitopes (a process known as "deimmunization"), using for example the methods disclosed in WO 98/52976 or WO 00/34317. Briefly, the heavy and light chain variable domains of an antibody can be analyzed for peptides that bind MHC class II; these peptides represent potential T cell epitopes (as defined in WO 98/52976 and WO 00/34317). To detect potential T cell epitopes, a computer modeling method known as "peptide threading" can be applied and additionally the database of human MHC class II binding peptides can be searched for motifs present in the VH and VL sequences as described in WO 98/52976 and WO 00/34317. These motifs bind to any of the 18 major MHC class II DR allotypes and thus constitute potential T cell epitopes. The potential T cell epitopes detected can be eliminated by substituting a small number of amino acid residues in the variable domain or preferably by single amino acid substitutions. Typically, conservative substitutions are made. Frequently, but not exclusively, amino acids that are common at positions in human germline antibody sequences are used. Human germline sequences are disclosed, for example, in Tomlinson et al (1992) j.mol.biol.227: 776-798; cook, g.p. et al (1995) immunol.today vol.16 (5): 237-242; and Tomlinson et al (1995) EMBO J.14: 14: 4628 and 4638. The VBASE catalog provides a comprehensive catalog of human immunoglobulin variable region sequences (codified by Tomlinson, LA. et al, MRC Centre for Protein Engineering, Cambridge, UK). These sequences can be used as a source of human sequences, e.g., for framework regions and CDRs. Common human framework regions may also be used, for example as described in U.S. Pat. No. 6,300,064.

A "humanized" antibody, antibody construct, variant or fragment thereof (e.g., Fv, Fab ', F (ab') 2, or other antigen-binding subsequence of an antibody) is an antibody or immunoglobulin of predominantly human sequence that contains minimal sequence from a non-human immunoglobulin. For the most part, humanized antibodies are human immunoglobulins (recipient antibody) in which residues from a hypervariable region (and also the CDRs) of the recipient are replaced by residues from a hypervariable region of a non-human (e.g. rodent) species (donor antibody) such as mouse, rat, hamster, or rabbit having the desired specificity, affinity, and capacity. In some cases, Fv Framework Region (FR) residues of the human immunoglobulin are replaced with corresponding non-human residues. In addition, "humanized antibodies" as used herein may also comprise residues not found in recipient antibodies and donor antibodies. These modifications were made to further improve and optimize antibody performance. The humanized antibody may also comprise at least a portion of an immunoglobulin constant region (Fc), typically that of a human immunoglobulin. For more details, see Jones et al, Nature, 321: 522-525 (1986); reichmann et al, Nature, 332: 323-329 (1988); and Presta, curr.op.struct.biol., 2: 593-596(1992).

Humanized antibodies or fragments thereof can be produced by replacing the sequence of an Fv variable domain that is not directly involved in antigen binding with an equivalent sequence from a human Fv variable domain. An exemplary method for producing a humanized antibody or fragment thereof is described by Morrison (1985) Science 229: 1202-1207; oi et al (1986) BioTechniques 4: 214; and US 5,585,089, US 5,693,761, US 5,693,762, US 5,859,205, and US 6,407,213. These methods include isolating, manipulating and expressing nucleic acid sequences encoding all or part of an immunoglobulin Fv variable domain from at least one of the heavy or light chains. Such nucleic acids can be obtained from hybridomas that produce antibodies to a predetermined target (as described above), as well as from other sources. The recombinant DNA encoding the humanized antibody molecule can then be cloned into an appropriate expression vector.

Humanized antibodies can also be produced using transgenic animals such as mice that express human heavy and light chain genes but are incapable of expressing endogenous mouse immunoglobulin heavy and light chain genes. Winter describes an exemplary CDR grafting method that can be used to prepare the humanized antibodies described herein (U.S. patent No. 5,225,539). All CDRs of a particular human antibody can be replaced by at least a portion of the non-human CDRs, or only some CDRs can be replaced by non-human CDRs. Only the number of CDRs required to bind the humanized antibody to the predetermined antigen need be replaced.

Humanized antibodies can be optimized by introducing conservative substitutions, consensus sequence substitutions, germline substitutions and/or back mutations. Such altered immunoglobulin molecules can be prepared by any of several techniques known in the art (e.g., Teng et al, Proc. Natl. Acad. Sci. U.S.A., 80: 7308-7312, 1983; Kozbor et al, Immunology Today, 4: 7279, 1983; Olsson et al, meth. enzymol., 92: 3-16, 1982 and EP 239400).

The terms "human antibody", "human antibody construct" and "human binding domain" include antibodies, antibody constructs and binding domains having antibody regions (e.g., variable and constant regions or variable and constant domains substantially corresponding to human germline immunoglobulin sequences known in the art), including, for example, those described by Kabat et al (1991) (supra). The human antibodies, antibody constructs or binding domains of the invention can include amino acid residues that are not encoded by human germline immunoglobulin sequences, for example, in the CDRs, particularly in CDR3 (e.g., mutations introduced by random or site-specific mutagenesis in vitro or by somatic mutation in vivo). The human antibody, antibody construct or binding domain may have at least one, two, three, four, five or more positions substituted with amino acid residues not encoded by human germline immunoglobulin sequences. The definition of human antibodies, antibody constructs and binding domains as used herein also contemplates fully human antibodies, which include only non-artificial and/or genetically altered human antibody sequences, such as antibody sequences that can be derived by using techniques or systems such as xenolouse.

In some embodiments, the antibody constructs of the invention are "isolated" or "substantially pure" antibody constructs. When used to describe the antibody constructs disclosed herein, "isolated" or "substantially pure" refers to an antibody construct that is identified, isolated, and/or recovered from a component of its production environment. Preferably, the antibody construct does not bind or does not substantially bind to all other components from its production environment. Contaminant components of the environment in which they are produced (e.g., contaminants derived from recombinantly transfected cells) are substances that would normally interfere with diagnostic or therapeutic uses of the polypeptide, and may include enzymes, hormones, and other proteinaceous or nonproteinaceous solutes. The antibody construct may, for example, constitute at least about 5 wt%, or at least about 50 wt%, of the total protein in a given sample. It is understood that the isolated protein may represent from 5 wt% to 99.9 wt% of the total protein content, as the case may be. By using inducible promoters or high expression promoters, polypeptides can be produced with significantly higher concentrations, thereby producing polypeptides at increased concentration levels. This definition includes the production of antibody constructs in a variety of organisms and/or host cells known in the art. In preferred embodiments, the antibody construct will be (1) purified by using a rotor sequencer to an extent sufficient to obtain at least 15 residues of the N-terminal or internal amino acid sequence, or (2) purified by SDS-PAGE under non-reducing or reducing conditions to homogeneity using coomassie blue or preferably silver stain. However, typically an isolated antibody construct will be prepared by at least one purification step.

According to the present invention, the term "binding domain" characterizes a domain that (specifically) binds to/interacts with or recognizes a given target epitope or a given target site on a target molecule (antigen, here EGFRVIII and CD3, respectively). The structure and function of the first binding domain (recognizing EGFRVIII), and preferably also the structure and/or function of the second binding domain (recognizing CD3) is based on the structure and/or function of an antibody, e.g. a full-length or intact immunoglobulin molecule. According to the invention, the first binding domain is characterized by the presence of three light chain CDRs (i.e. CDR1, CDR2 and CDR3 of the VL region) and/or three heavy chain CDRs (i.e. CDR1, CDR2 and CDR3 of the VH region). The second binding domain preferably also comprises antibody minimum structural requirements that allow target binding. More preferably, the second binding domain comprises at least three light chain CDRs (i.e., CDR1, CDR2, and CDR3 of the VL region) and/or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 of the VH region). It is envisaged that the first and/or second binding domains are generated or obtainable from phage display or library screening methods, rather than being generated or obtained by grafting CDR sequences from a pre-existing (monoclonal) antibody into a scaffold.

According to the invention, the binding domain is in the form of one or more polypeptides. Such polypeptides may include a proteinaceous moiety and a non-proteinaceous moiety (e.g., a chemical linker or a chemical cross-linking agent such as glutaraldehyde). Proteins (including fragments thereof, preferably biologically active fragments and peptides typically having less than 30 amino acids) comprise two or more amino acids (resulting in amino acid chains) coupled to each other via covalent peptide bonds. The term "polypeptide" as used herein describes a group of molecules typically consisting of more than 30 amino acids. The polypeptides may further form multimers, such as dimers, trimers and higher oligomers, i.e. consisting of more than one polypeptide molecule. The polypeptide molecules forming such dimers, trimers, etc. may be the same or different. Thus, the corresponding higher order structures of such multimers are referred to as homo-or heterodimers, homo-or heterotrimers, and the like. An example of a heteromultimer is an antibody molecule, which in its naturally occurring form consists of two identical light polypeptide chains and two identical heavy polypeptide chains. The terms "peptide", "polypeptide" and "protein" also refer to naturally modified peptides/polypeptides/proteins, wherein the modification is effected, for example, by post-translational modifications such as glycosylation, acetylation, phosphorylation, and the like. When referred to herein, a "peptide", "polypeptide" or "protein" may also be chemically modified, e.g., pegylated. Such modifications are well known in the art and are described below.

Preferably, the binding domain that binds to EGFRVIII and/or the binding domain that binds to CD3 is a human binding domain. Antibodies and antibody constructs comprising at least one human binding domain avoid some of the problems associated with antibodies or antibody constructs having non-human (e.g., rodent (e.g., mouse, rat, hamster, or rabbit)) variable and/or constant regions. The presence of such rodent derived proteins may result in rapid clearance of the antibody or antibody construct, or may result in the generation of an immune response against the antibody or antibody construct by the patient. To avoid the use of rodent derived antibodies or antibody constructs, human or fully human antibody/antibody constructs can be produced by introducing human antibody functions into rodents such that they produce fully human antibodies.

The ability to clone and reconstitute megabase-sized human loci in YACs and introduce them into the mouse germline provides a powerful method of elucidating the functional components of very large or roughly mapped loci and generating useful models of human disease. Moreover, the replacement of the mouse locus with its human equivalent using this technique can provide a unique understanding of the expression and regulation of human gene products during development, communication with other systems, and their involvement in disease induction and progression.

An important practical application of this strategy is the "humanization" of the mouse humoral immune system. The introduction of human immunoglobulin (Ig) loci into mice where endogenous Ig genes have been inactivated provides an opportunity to study the potential mechanisms of programmed expression and assembly of antibodies and their role in B cell development. Furthermore, this strategy can provide an ideal source for the production of fully human monoclonal antibodies (mabs), thereby becoming an important milestone in the prospect of antibody therapy for achieving human disease. Fully human antibodies or antibody constructs are expected to minimize the immunogenic and allergic reactions inherent to mouse mabs or mabs of mouse origin, thereby increasing the efficacy and safety of the administered antibody/antibody construct. The use of fully human antibodies or antibody constructs is expected to provide great advantages in the treatment of chronic and recurrent human diseases (e.g., inflammation, autoimmunity, and cancer) that require repeated compound administration.

One way to achieve this goal is to engineer mouse strains with defects in mouse antibody production with large segments of human Ig loci, which would be expected to produce large repertoires of human antibodies in the absence of mouse antibodies. Large human Ig fragments will retain large variable gene diversity and proper regulation of antibody production and expression. By exploiting the mouse mechanism to achieve antibody diversification and selection and lack of immune tolerance to human proteins, the repertoire of human antibodies re-generated in these mouse strains should produce high affinity antibodies to any antigen of interest, including human antigens. Using hybridoma technology, antigen-specific human mabs with the desired specificity can be readily generated and selected. This general strategy was demonstrated in conjunction with the generation of the first generation XenoMouse mouse strain (see Green et al, Nature Genetics 7: 13-21 (1994)). The XenoMouse line was engineered with Yeast Artificial Chromosomes (YACs) containing germline-configured fragments of 245kb and 190kb size containing core variable and constant region sequences for the human heavy and kappa light chain loci, respectively. YACs containing human igs were shown to be compatible with mouse systems in antibody rearrangement and expression, and could replace inactivated mouse Ig genes. This is manifested by its ability to induce B cell development, generate an adult-like repertoire of fully human antibodies, and generate antigen-specific human mabs. These results also indicate that the introduction of a larger portion of the human Ig locus containing a greater number of V genes, additional regulatory elements and human Ig constant regions may recapitulate a substantially complete repertoire featuring a human humoral response to infection and immunity. The work of Green et al has recently been expanded to introduce a repertoire of human antibodies greater than about 80% by introducing megabase-sized germline configuration YAC fragments of the human heavy chain locus and kappa light chain locus, respectively. See Mendez et al, Nature Genetics 15: 146- & 156(1997) and U.S. patent application No. 08/759,620.

The generation of XenoMouse mice is further discussed and described in the following references: U.S. patent application serial No. 07/466,008, serial No. 07/610,515, serial No. 07/919,297, serial No. 07/922,649, serial No. 08/031,801, serial No. 08/112,848, serial No. 08/234,145, serial No. 08/376,279, serial No. 08/430,938, serial No. 08/464,584, serial No. 08/464,582, serial No. 08/463,191, serial No. 08/462,837, serial No. 08/486,853, serial No. 08/486,857, serial No. 08/486,859, serial No. 08/462,513, serial No. 08/724,752, and serial No. 08/759,620; and U.S. Pat. nos. 6,162,963, 6,150,584, 6,114,598, 6,075,181 and 5,939,598, and japanese patent nos. 3068180B 2, 3068506B 2 and 3068507B 2. See also Mendez et al, Nature Genetics 15: 146-: 483-495(1998), EP 0463151B 1, WO 94/02602, WO 96/34096, WO 98/24893, WO 00/76310 and WO 03/47336.

Among the alternatives, other companies, including Genpharm International, Inc., have adopted the "minilocus" approach. In the minilocus approach, foreign Ig loci are mimicked by the inclusion of fragments (single genes) from the Ig locus. Thus, one or more VH genes, one or more DH genes, one or more JH genes, a μ constant region, and a second constant region (preferably a γ constant region) are formed into constructs for insertion into an animal. Such a process is described in U.S. patent No. 5,545,807 to Surani et al; U.S. Pat. Nos. 5,545,806, 5,625,825, 5,625,126, 5,633,425, 5,661,016, 5,770,429, 5,789,650, 5,814,318, 5,877,397, 5,874,299 and 6,255,458 to Lonberg and Kay et al; U.S. patent nos. 5,591,669 and 6,023,010 to Krimpenfort and Berns et al; U.S. Pat. Nos. 5,612,205, 5,721,367, and 5,789,215 to Berns et al; choi and Dunn et al, U.S. patent No. 5,643,763; and U.S. patent application Ser. No. 07/574,748, Ser. No. 07/575,962, Ser. No. 07/810,279, Ser. No. 07/853,408, Ser. No. 07/904,068, Ser. No. 07/990,860, Ser. No. 08/053,131, Ser. No. 08/096,762, Ser. No. 08/155,301, Ser. No. 08/161,739, Ser. No. 08/165,699, Ser. No. 08/209,741, to Genpharm International. See also EP 0546073B 1, WO 92/03918, WO 92/22645, WO 92/22647, WO 92/22670, WO 93/12227, WO 94/00569, WO 94/25585, WO 96/14436, WO 97/13852 and WO 98/24884 and us patent No. 5,981,175. See further Taylor et al (1992), Chen et al (1993), Tuaillon et al (1993), Choi et al (1993), Lonberg et al (1994), Taylor et al (1994), Tuaillon et al (1995), Fishwild et al (1996).

Kirin also describes the production of human antibodies from mice in which large segments of chromosomes or entire chromosomes are introduced by minicell fusion. See european patent application nos. 773288 and 843961. Xenrex Biosciences are developing a technology for potential production of human antibodies. In this technique, SCID mice are reconstituted with human lymphocytes such as B and/or T cells. The mouse is then immunized with the antigen and an immune response can be raised to the antigen. See U.S. patent nos. 5,476,996, 5,698,767, and 5,958,765.

Human anti-mouse antibody (HAMA) responses have prompted the industry to begin making chimeric or otherwise humanized antibodies. However, it is expected that certain human anti-chimeric antibody (HACA) responses will be observed, particularly in chronic or multi-dose use of the antibody. Accordingly, it would be desirable to provide antibody constructs comprising a human binding domain to EGFRVIII and a human binding domain to CD3, in order to eliminate the concerns and/or effects of HAMA or HACA responses.

According to the present invention, the terms "(specific) binding", "(specific) recognition", "(specific) targeting" and "reacting with" (specific) mean that the binding domain interacts or specifically interacts with a given epitope or a given target site on the target molecule (antigen, here EGFRVIII and CD3, respectively).

The term "epitope" refers to a site on an antigen to which a binding domain (e.g., an antibody or immunoglobulin, or a derivative, fragment, or variant of an antibody or immunoglobulin) specifically binds. An "epitope" is antigenic, and thus the term epitope is also sometimes referred to herein as an "antigenic structure" or "antigenic determinant. Thus, the binding domain is the "antigen-interaction-site". The binding/interaction is also understood to define "specific recognition".

An "epitope" may be formed of contiguous amino acids or noncontiguous amino acids juxtaposed by tertiary folding of a protein. A "linear epitope" is an epitope in which the primary sequence of amino acids comprises the recognized epitope. Linear epitopes typically comprise at least 3 or at least 4, more typically at least 5 or at least 6 or at least 7, for example from about 8 to about 10 amino acids in the unique sequence.

In contrast to linear epitopes, a "conformational epitope" is an epitope in which the primary sequence of amino acids comprising the epitope is not the only defining component of the epitope being recognized (e.g., an epitope in which the primary sequence of amino acids is not necessarily recognized by the binding domain). Typically, conformational epitopes comprise an increased number of amino acids relative to linear epitopes. With respect to the recognition of conformational epitopes, the binding domains recognize the three-dimensional structure of an antigen, preferably a peptide or protein or fragment thereof (in the context of the present invention, the antigenic structure directed to one of the binding domains is comprised within the EGFRVIII protein). For example, when a protein molecule is folded to form a three-dimensional structure, certain amino acids and/or polypeptide backbones that form the conformational epitope become juxtaposed, enabling the antibody to recognize the epitope. Methods for determining epitope conformation include, but are not limited to, x-ray crystallography, two-dimensional nuclear magnetic resonance (2D-NMR) spectroscopy, and site-specific spin labeling, as well as Electron Paramagnetic Resonance (EPR) spectroscopy.

One method for epitope mapping is described below: when regions (contiguous amino acid extensions) in the human EGFRVIII protein are exchanged/replaced with corresponding regions of non-human and non-primate EGFRVIII antigens (e.g., mouse EGFRVIII, but other animals such as chickens, rats, hamsters, rabbits, etc. are also conceivable), reduced binding of the binding domain is expected unless the binding domain is cross-reactive to the non-human, non-primate EGFRVIII used. The reduction is preferably at least 10%, 20%, 30%, 40% or 50% compared to binding to a corresponding region in a human EGFRVIII protein (wherein binding to a corresponding region in a human EGFRVIII protein is set to 100%); more preferably at least 60%, 70% or 80%, and most preferably 90%, 95% or even 100%.

Another method of determining the contribution of specific residues of a target antigen to the recognition of an antibody construct or binding domain is alanine scanning (see, e.g., Morrison KL and Weiss GA, Cur Opin Chem biol.2001 Jun; 5 (3): 302-7), in which each residue to be analyzed is replaced by alanine, e.g., by site-directed mutagenesis. Alanine is used because it has a non-bulky, chemically inert methyl functionality, but still mimics the secondary structure reference possessed by many other amino acids. Sometimes where it is desired to maintain the size of the mutated residue, bulky amino acids such as valine or leucine may be used. Alanine scanning is a well-established technique that has been used for a long time.

The interaction between the binding domain and the epitope or epitope-containing region means that the binding domain exhibits appreciable affinity for the epitope/epitope-containing region on a particular protein or antigen (here EGFRVIII and CD3, respectively) and generally does not exhibit significant reactivity with proteins or antigens other than EGFRVIII or CD 3. "appreciable affinity" includes at least about 10^-6M (KD) or stronger affinity binding. Preferably, when the binding affinity is about 10^-12To 10^-8M、10^-12To 10^-9M、10^-12To 10^-10M、10^-11To 10^-8M, preferably about 10^-11To 10^-9M, binding is considered specific. Whether or not the binding domain specifically reacts or binds to the target, testing can be readily performed, particularly by comparing the reaction of the binding domain and the target protein or antigen to the reaction of the binding domain and a protein or antigen other than EGFRVIII or CD 3. Preferably, the binding domains of the invention do not substantially or essentially bind to proteins or antigens other than EGFRVIII or CD3 (i.e., the first binding domain is not capable of binding to proteins other than EGFRVIII and the second binding domain is not capable of binding to proteins other than CD 3).

The term "does not substantially/essentially bind" or "does not bind" means that the binding domain of the invention does not bind to proteins or antigens other than EGFRVIII or CD3, i.e. does not show more than 30%, preferably not more than 20%, more preferably not more than 10%, particularly preferably not more than 9%, 8%, 7%, 6% or 5% reactivity with proteins or antigens other than EGFRVIII or CD3, wherein binding to EGFRVIII or CD3, respectively, is set to 100%.

Specific binding is believed to be achieved by specific motifs in the amino acid sequences of the binding domains and the antigen. Thus, binding is achieved due to their primary, secondary and/or tertiary structure and secondary modifications of the structure. Specific interaction of an antigen-interaction-site with its specific antigen can result in simple binding of the site to the antigen. Furthermore, the specific interaction of an antigen-interaction-site with its specific antigen may alternatively or additionally result in the priming of a signal, e.g. due to induction of a change in antigen conformation, oligomerization of the antigen, etc.

The epitope of the bispecific antibody constructs of the invention is located at the EGFRvIII-specific junction between amino acid residues 5 and 274 of EGFR, resulting from a mutation of EGFR into a splicing mutation EGFRvIII. This mutation removes residues 2-273 of the mature EGFR sequence while introducing a single glycine residue (see fig. 1).

In an embodiment of the invention it is further preferred that the second binding domain binds to human CD3 epsilon and to common marmoset (Callithrix jacchus), tamarisk villous (Saguinus Oedipus) or squirrel (Saimiri sciureus) CD3 epsilon.

The term "variable" refers to portions of an antibody or immunoglobulin domain that exhibit sequence variability and relate to the portions that determine the specificity and binding affinity of a particular antibody (i.e., the "variable domain"). The pairing of the variable heavy chain (VH) and variable light chain (VL) together forms a single antigen binding site.

Variability is not evenly distributed throughout the variable region of the antibody; it is concentrated in the respective subdomains of the heavy and light chain variable regions. These subdomains are referred to as "hypervariable regions" or "complementarity determining regions" (CDRs). The more conserved (i.e., non-hypervariable) portion of the variable domain is called the "framework" region (FRM or FR) and provides a scaffold of six CDRs in three-dimensional space to form the antigen-binding surface. The variable domains of naturally occurring heavy and light chains each comprise four FRM regions (FR1, FR2, FR3 and FR4) which adopt predominantly a β -sheet configuration and are linked by three hypervariable regions which form loops connecting, and in some cases forming part of, the β -sheet structure. The hypervariable regions in each chain are tightly joined together by the FRM and, together with the hypervariable regions from the other chain, facilitate the formation of an antigen-binding site (see Kabat et al, supra).

The term "CDR" refers to the complementarity determining regions three of which constitute the binding characteristics of the light chain variable region (CDR-L1, CDR-L2 and CDR-L3) and three of which constitute the binding characteristics of the heavy chain variable region (CDR-H1, CDR-H2 and CDR-H3). The CDRs contain most of the residues responsible for the specific interaction of the antibody with the antigen and thus contribute to the functional activity of the antibody molecule: they are the major determinants of antigen specificity.

The CDR boundaries and lengths on the exact definition are limited by different classification and numbering systems. Accordingly, CDRs may be referenced by Kabat, Chothia, contact, or any other boundary definition (including the numbering systems described herein). Although having different boundaries, these systems each have some degree of overlap in the portions that make up the so-called "hypervariable regions" within the variable sequences. The CDR definitions according to these systems may thus differ in length and boundary region with respect to adjacent framework regions. See, e.g., Kabat (a method based on sequence variability across species), Chothia (a method based on crystallographic studies of antigen-antibody complexes) and/or MacCallum (Kabat et al, supra; Chothia et al, J.mol.biol, 1987, 196: 901-. Another criterion for characterizing antigen binding sites is the AbM definition used by Oxford Molecular's AbM antibody modeling software. See, for example, Protein Sequence and Structure Analysis of Antibody Variable Domains, as found in Antibody Engineering Lab Manual (ed.: Duebel, S. and Kontermann, R., Springer-Verlag, Heidelberg). To the extent that both residue identification techniques define overlapping regions rather than identical regions, they can be combined to define hybrid CDRs. However, the numbering is preferably according to the so-called Kabat system.

In general, CDRs form a loop structure that can be classified as canonical structures. The term "canonical structure" refers to the backbone conformation adopted by the antigen binding (CDR) loops. Five of the six antigen-binding loops have been found from comparative structural studies to have only a limited set of available conformations for the library. Each canonical structure can be characterized by the twist angle of the polypeptide backbone. Thus, corresponding loops between antibodies can have very similar three-dimensional structures, although most of the loops have a high degree of amino acid sequence variability (Chothia and Lesk, J.mol.biol., 1987, 196: 901; Chothia et al, Nature, 1989, 342: 877; Martin and Thornton, J.mol.biol., 1996, 263: 800). In addition, there is a link between the loop structure employed and the amino acid sequence surrounding it. The conformation of a particular canonical class is determined by the length of the loop and the amino acid residues at key positions within the loop and within the conserved framework (i.e., outside the loop). Thus, assignment to a particular canonical class can be based on the presence of these key amino acid residues.

The term "canonical structure" may also include considerations regarding the linear sequence of antibodies, e.g., the Kabat classification (Kabat et al, supra). The Kabat numbering scheme (system) is a widely adopted standard for numbering amino acid residues of antibody variable domains in a consistent manner, and is also a preferred scheme for use in the present invention, as described elsewhere herein. Additional structural considerations may also be used to determine the canonical structure of the antibody. For example, those differences that are not adequately reflected by Kabat numbering can be described using the numbering system of Chothia et al and/or revealed by other techniques (e.g., crystallography and two-or three-dimensional computer modeling). Thus, a given antibody sequence can be assigned to a canonical class that allows, among other things, the identification of the appropriate base sequence (e.g., based on the desire to include various canonical structures in the library). Kabat numbering of antibody amino acid sequences and structural considerations as described by Chothia et al (supra) and their effect on the specification of the interpretation of antibody structures are described in the literature. The subunit structures and three-dimensional configurations of different classes of immunoglobulins are well known in the art. For a review of antibody structures, see Antibodies: a Laboratory Manual, Cold Spring Harbor Laboratory, edited by Harlow et al, 1988.

The CDR3 of the light chain, particularly CDR3 of the heavy chain, may constitute the most important determinant in antigen binding within the variable regions of the light and heavy chains. In some antibody constructs, the heavy chain CDR3 appears to constitute the primary contact region between the antigen and the antibody. In vitro options where CDR3 is altered alone can be used to alter the binding characteristics of the antibody or to determine which residues contribute to antigen binding. Thus, CDR3 is generally the largest source of molecular diversity within an antibody binding site. For example, H3 may be as short as two amino acid residues or more than 26 amino acids.

In a classical full-length antibody or immunoglobulin, each light chain (L) is linked to a heavy chain (H) by one covalent disulfide bond, while the two H chains are linked to each other by one or more disulfide bonds (depending on the H chain isotype). The CH domain closest to VH is commonly designated CH 1. The constant ("C") domain is not directly involved in antigen binding, but exhibits various effector functions, such as antibody-dependent cell-mediated cytotoxicity and complement activation. The Fc region of an antibody is contained within the heavy chain constant domain and is, for example, capable of interacting with Fc receptors located on the cell surface.

The sequences of the antibody genes after assembly and somatic mutation were highly variable, and it was estimated that these varied genes encode 10¹⁰A variety of antibody molecules (Immunoglobulin Genes, 2 nd edition, edited by Jonio et al, Academic Press, San Diego, Calif., 1995). Thus, the immune system provides a repertoire of immunoglobulins. The term "repertoire" refers to at least one nucleotide sequence derived, in whole or in part, from at least one sequence encoding at least one immunoglobulin. The sequences may be generated by in vivo rearrangement of V, D and J segments of the heavy chain and V and J segments of the light chain. Alternatively, the sequences may be produced from cells that undergo rearrangement in response to, for example, in vitro stimulation. Alternatively, part or all of the sequence may be obtained by DNA splicing, nucleotide synthesis, mutagenesis, and other methods, see, e.g., U.S. patent No. 5,565,332. Group library canIncluding only one sequence or may include multiple sequences, including sequences in a collection of genetic diversity.

As used herein, the term "bispecific" refers to an "at least bispecific" antibody construct, i.e. said antibody construct comprises at least a first binding domain and a second binding domain, wherein the first binding domain binds to one antigen or target (here EGFRVIII) and the second binding domain binds to another antigen or target (here CD 3). Thus, the antibody construct according to the invention comprises specificity for at least two different antigens or targets. The term "bispecific antibody construct" of the present invention also includes multispecific antibody constructs such as trispecific antibody constructs comprising three binding domains, or constructs having more than three (e.g., four, five) specificities.

Given that the antibody constructs according to the invention are (at least) bispecific, they are not naturally occurring and differ significantly from naturally occurring products. Thus, a "bispecific" antibody construct or immunoglobulin is an artificial hybrid antibody or immunoglobulin comprising at least two different binding sites with different specificities. Bispecific antibody constructs can be produced by a variety of methods, including fusion of hybridomas or ligation of Fab' fragments. See, e.g., Songsivilai & Lachmann, clin. exp. immunol.79: 315-321(1990).

The at least two binding and variable domains of the antibody construct of the invention may or may not comprise a peptide linker (spacer peptide). According to the present invention, the term "peptide linker" comprises an amino acid sequence linking the amino acid sequences of one (variable and/or binding) domain and the other (variable and/or binding) domain of the antibody construct of the invention to each other. The essential technical feature of this peptide linker is that it does not contain any polymerization activity. Among suitable peptide linkers are those described in U.S. Pat. Nos. 4,751,180 and 4,935,233 or WO 88/09344. Peptide linkers can also be used to attach other domains or moieties or regions (such as half-life extension domains) to the antibody constructs of the invention.

In the case of using a joint, theThe linker preferably has a length and sequence sufficient to ensure that each of the first and second domains is capable of maintaining their differential binding specificities independently of each other. For peptide linkers connecting at least two binding domains (or two variable domains) in the antibody construct of the invention, peptide linkers comprising only a few amino acid residues (e.g. 12 amino acid residues or less) are preferred. Thus, peptide linkers having 12, 11, 10, 9, 8,7, 6 or 5 amino acid residues are preferred. Contemplated peptide linkers having less than 5 amino acids comprise 4,3, 2 or 1 amino acids, with Gly-rich linkers being preferred. In the context of said "peptide linker", a particularly preferred "single" amino acid is Gly. Thus, the peptide linker may consist of a single amino acid Gly. Another preferred embodiment of the peptide linker is characterized by the amino acid sequence Gly-Gly-Gly-Gly-Ser, i.e.Gly₄Ser (SEQ ID NO: 1) or a polymer thereof, i.e. (Gly)₄Ser) x, wherein x is an integer of 1 or more (e.g., 2 or 3). Preferred linkers are shown in SEQ ID NO: 1-9. The characteristics of the peptide linker including the absence of secondary structure promoting factors are known in the art and are described, for example, in Dall' Acqua et al (Biochem. (1998)37, 9266-9273); cheadle et al (Mol Immunol (1992)29, 21-30) and Raag and Whitlow (FASEB (1995)9(1), 73-80). Peptide linkers that do not otherwise promote any secondary structure are preferred. The linking of the domains to each other can be achieved, for example, by genetic engineering, as described in the examples. Methods for making fused and operably linked bispecific single chain constructs and expressing the constructs in mammalian cells or bacteria are well known in the art (e.g., WO 99/54440 or Sambrook et al, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 2001).

As noted above, the present invention provides preferred embodiments wherein the antibody construct is in a form selected from the group consisting of: (scFv)₂scFv-single domain mAbs, diabodies and oligomers of any of these forms.

According to a particularly preferred embodiment, and as described in the accompanying examples, the antibody construct of the invention is a "bispecific single chain antibody construct", more preferably a bispecific "single chain Fv" (scFv). Although the two domains of the Fv fragment, VL and VH, are encoded by different genes, they can be joined using recombinant methods by synthetic linkers (as described above) such that they can be made as a single protein chain in which the VL and VH regions pair to form monovalent molecules; see, e.g., Huston et al (1988) Proc.Natl.Acad.Sci USA 85: 5879-5883). These antibody fragments are obtained using conventional techniques known to those skilled in the art and the function of the fragments is assessed in the same manner as for intact or full-length antibodies. Single chain variable fragments (scFv) are thus fusion proteins of the heavy (VH) and light (VL) chain variable regions of an immunoglobulin, typically linked with a short linker peptide of about 10 to about 25 amino acids, preferably about 15 to 20 amino acids. The linker is typically glycine-rich for flexibility and serine or threonine for solubility and can link the N-terminus of VH with the C-terminus of VL, or vice versa. This protein retains the specificity of the original immunoglobulin despite the removal of the constant region and the introduction of the linker.

Bispecific single chain molecules are known in the art and described in WO 99/54440; mack, j.immunol. (1997), 158, 3965-; mack, PNAS, (1995), 92, 7021-; kufer, Cancer immunol., (1997), 45, 193-197;

blood, (2000), 95, 6, 2098 and 2103; bruhl, Immunol., (2001), 166, 2420-; kipriyanov, J.mol.biol., (1999), 293, 41-56. The techniques described for the production of single chain antibodies (see, inter alia, U.S. Pat. No. 4,946,778; Kontermann and Dubel (2010), supra; and Little (2009), supra) may be adapted to produce single chain antibody constructs that specifically recognize a selected target.

Bivalent (also called bivalent) or bispecific single chain variable fragments (with (scFv)₂The form bi-scFvs or di-scFvs) may be engineered by linking two scFv molecules (e.g. using a linker as described above). If the two scFv are dividedThe same binding specificity, the resulting (scFv)₂The molecule will preferably be referred to as bivalent (i.e. it has two valencies against the same target epitope). If two scFv molecules have different binding specificities, the result is (scFv)₂The molecule will preferably be referred to as bispecific. Ligation can be accomplished by generating a single peptide chain with two VH regions and two VL regions to generate a tandem scFv (see, e.g., Kufer P. et al, (2004) Trends in Biotechnology 22 (5): 238-244). Another possibility is to generate scFv molecules with a peptide linker that is too short (e.g., about 5 amino acids) for the two variable regions to fold together, thereby forcing scFv dimerization. This type is known as diabodies (see, e.g., Hollinger, Philipp et al, (7. 1993) Proceedings of the National Academy of Sciences of the United States of America 90 (14): 6444-8.).

According to another preferred embodiment of the antibody construct of the invention, the heavy chain (VH) and the light chain (VL) of the binding domain that binds to the target antigen EGFRVIII or CD3 are not directly linked via the above-mentioned peptide linker, but form the binding domain as a result of the formation of a bispecific molecule as described above for diabodies. Thus, the VH chain of the CD3 binding domain may be fused to the VL of the EGFRVIII binding domain via such a peptide linker, whereas the VH chain of the EGFRVIII binding domain is fused to the VL of the CD3 binding domain via such a peptide linker.

Single domain antibodies comprise only one (monomeric) antibody variable domain, which is capable of selectively binding a particular antigen independently of other V regions or domains. The first single domain antibodies were engineered from heavy chain antibodies found in camelids and these antibodies were designated V_HAnd (e) H fragment. Cartilaginous fish also have heavy chain antibodies (IgNAR) from which they can be obtained called V_NARA single domain antibody of a fragment. An alternative approach is to break down the dimeric variable domains of common immunoglobulins (e.g. from humans or rodents) into monomers, thereby obtaining VH or VL as single domain Ab. While most current research on single domain antibodies is based on heavy chain variable domains, nanobodies derived from light chains also exhibit affinity to targetSite specific binding. Examples of single domain antibodies are referred to as sdabs, nanobodies, or single variable domain antibodies.

Thus, (Single Domain mAb)₂Is composed of (at least) two single domain monoclonal antibodies independently selected from the group comprising VH, VL, V_HH and V_NARGroup) of the monoclonal antibody construct. The linker is preferably in the form of a peptide linker. Similarly, a "scFv-single domain mAb" is a monoclonal antibody construct consisting of at least one single domain antibody as described above and one scFv molecule as described above. Also, the linker is preferably in the form of a peptide linker.

T cells or T lymphocytes are a class of lymphocytes (which are themselves leukocytes) that play a central role in cell-mediated immunity. There are several subsets of T cells, each with a unique function. T cells can be distinguished from other lymphocytes such as B cells and NK cells by the presence of a T Cell Receptor (TCR) on the cell surface. The TCR is responsible for recognizing antigens bound to Major Histocompatibility Complex (MHC) molecules and is composed of two distinct protein chains. In 95% of T cells, the TCR is composed of α (alpha) and β (beta) chains. When TCRs are conjugated to antigenic peptides and MHC (peptide/MHC complex), T lymphocytes are activated by a series of biochemical events mediated by associated enzymes, co-receptors, specialized adaptor molecules and activated or released transcription factors.

The CD3 receptor complex is a protein complex and consists of four chains. In mammals, the complex contains a CD3 γ (gamma) chain, a CD3 δ (delta) chain, and two CD3 epsilon (epsilon) chains. These chains associate with the T Cell Receptor (TCR) and the so-called zeta (zeta) chain, forming the T cell receptor CD3 complex, and generate an activation signal in T lymphocytes. The CD3 γ (gamma), CD3 δ (delta), and CD3 epsilon (epsilon) chains are highly related cell surface proteins of the immunoglobulin superfamily that contain a single extracellular immunoglobulin domain. The intracellular tail of the CD3 molecule contains a single conserved motif called the immunoreceptor tyrosine-based activation motif or simply ITAM, which is essential for the signaling ability of the TCR. The CD3 epsilon molecule is a polypeptide encoded in humans by the CD3E gene located on chromosome 11. The most preferred epitope for CD3 epsilon is contained within amino acid residues 1-27 of the extracellular domain of human CD3 epsilon.

The redirected lysis of target cells by recruitment of T cells by multispecific, at least bispecific antibody constructs involves cytolytic synapse formation and delivery of perforin and granzymes. The conjugated T cells are capable of continuous lysis of target cells and are not affected by immune escape mechanisms that interfere with peptide antigen processing and presentation or clonal T cell differentiation; see, for example, WO 2007/042261

Cytotoxicity mediated by the egfriiixcd 3 bispecific antibody construct can be measured in various ways. See example 1. The effector cells may be, for example, stimulated enriched (human) CD8 positive T cells or unstimulated (human) Peripheral Blood Mononuclear Cells (PBMCs). If the target cell is derived from, or expresses or is transfected with, macaque EGFRVIII, the effector cell should also be of macaque origin, e.g., a macaque T cell line, e.g., 4119 LnPx. The target cell should express EGFRVIII, e.g., human or cynomolgus EGFRVIII (at least its extracellular domain). The target cell may be a cell line (e.g., CHO) stably or transiently transfected with EGFRVIII (e.g., human or cynomolgus EGFRVIII). Alternatively, the target cell may be an EGFRVIII positive naturally expressing cell line, such as the human glioblastoma cell line U87 or DK-MG. For target cell lines expressing higher levels of EGFRVIII on the cell surface, EC50 values were expected to be generally lower. The ratio of effector cells to target cells (E: T) is typically about 10: 1, but may vary. The cytotoxic activity of the egfrviii cd3 bispecific antibody construct can be measured in a 51-chromium release assay (incubation time about 18 hours) or in a FACS-based cytotoxicity assay (incubation time about 48 hours). The incubation time of the assay (cytotoxic reaction) may also be modified. Other methods of measuring cytotoxicity are well known to the skilled artisan and include MTT or MTS assays; ATP-based assays, including bioluminescent assays; sulforhodamine b (srb) assay, WST assay, clonogenic assay, and ECIS technique.

The cytotoxic activity mediated by the egfriiixcd 3 bispecific antibody construct of the invention is preferably measured in a cell-based cytotoxicity assay. Also hasThis cytotoxic activity can be measured in a 51-chromium release assay. The cytotoxic activity is measured by EC₅₀Values are expressed as corresponding to half the maximum effective concentration (concentration of antibody construct inducing a cytotoxic response located midway between baseline and maximum). Preferably, the EC of the EGFRUIHxCD 3 bispecific antibody construct₅₀Values of 5000pM or 4000pM, more preferably 3000pM or 2000pM, even more preferably 1000pM or 500pM, even more preferably 400pM or 300pM, even more preferably 200pM, even more preferably 100pM, even more preferably 50pM, even more preferably 20pM or 10pM, and most preferably 5 pM.

Given above of EC₅₀The values may be measured in different assays. Those skilled in the art know that EC can be expected when using stimulated/enriched CD8+ T cells as effector cells compared to unstimulated PBMCs₅₀The value will be lower. Furthermore, it can be expected that EC will be achieved when the target cells express a high amount of target antigen, as compared with low-target-expression rats₅₀The value is lower. For example, EC of EGFRVIII xcd3 bispecific antibody construct when stimulated/enriched human CD8+ T cells are used as effector cells (and EGFRVIII transfected cells such as CHO cells or EGFRVIII positive human glioblastoma cell line U87 or DK-MG are used as target cells)₅₀The value is preferably 1000pM or less, more preferably 500pM or less, even more preferably 250pM or less, even more preferably 100pM or less, even more preferably 50pM or less, even more preferably 10pM or less, and most preferably 5pM or less. EC for EGFRUIHxCD 3 bispecific antibody constructs when human PBMC are used as effector cells₅₀The value is preferably ≦ 5000pM or ≦ 4000pM (particularly when the target cell is an EGFRVIII positive human glioblastoma cell line U87 or DK-MG), more preferably ≦ 2000pM (particularly when the target cell is an EGFRVIII transfected cell such as a CHO cell), more preferably ≦ 1000pM or ≦ 500pM, even more preferably ≦ 200pM, even more preferably ≦ 150pM, even more preferably ≦ 100pM, and most preferably ≦ 50pM or less. EC for EGFRUIXICXCD 3 bispecific antibody constructs when using a macaque T cell line such as LnPx4119 as the effector cell and a macaque EGFRVIIII transfected cell line such as CHO cell as the target cell line₅₀The values are preferably ≦ 2000pM or ≦ 1500pM, more preferably ≦ 1000pM or ≦ 500pM, even more preferably ≦ 300pM or ≦ 250pM, even more preferably ≦ 100pM, and most preferably ≦ 50 pM.

Preferably, the egfriiixcd 3 bispecific antibody construct of the invention does not induce/mediate or does not substantially induce/mediate lysis of EGFRVIII negative cells such as CHO cells. The terms "does not induce lysis", "essentially does not induce lysis", "does not mediate lysis" or "essentially does not mediate lysis" mean that the antibody construct of the invention does not induce or mediate lysis of more than 30%, preferably not more than 20%, more preferably not more than 10%, particularly preferably not more than 9%, 8%, 7%, 6% or 5% of EGFRVIII negative cells, whereby lysis of the EGFRVIII positive human glioblastoma cell line U87 or DK-MG (see above) is set to 100%. This is generally applicable to antibody constructs at concentrations up to 500 nM. The skilled person knows how to measure cell lysis without difficulty. Furthermore, the present specification teaches specific instructions how to measure cell lysis.

The difference in cytotoxic activity between the monomeric and dimeric isoforms of a single egfrviii xcd3 bispecific antibody construct is referred to as "potency difference". This difference in potency can be calculated, for example, as EC in monomeric and dimeric forms of the molecule₅₀See example 1.8 for ratios between values. The difference in potency of the EGFRVIII x CD3 bispecific antibody constructs of the invention is preferably 5 or less, more preferably 4 or less, even more preferably 3 or less, even more preferably 2 or less, further preferably 1 or less, and most preferably 0.3 or less.

The first and/or second (or any other) binding domain of the antibody construct of the invention preferably has cross-species specificity for a member of a primate mammal. For example, cross-species specific CD3 binding domains are described in WO 2008/119567. According to one embodiment, in addition to binding to human EGFRVIII and human CD3, respectively, the first and/or second binding domain will also bind to EGFRVIII/CD3 of primates including, but not limited to, new world primates (e.g., common marmosets, tamarisk or squirrel monkeys), old world primates (e.g., baboons and macaques), gibbons, orangutans and non-human subfamilies. It is envisaged that the first binding domain of the antibody construct of the invention that binds to human EGFRVIII on the surface of a target cell also binds at least cynomolgus monkey EGFRVIII, and/or the second binding domain that binds to human CD3 on the surface of a T cell also binds at least cynomolgus monkey CD 3. The preferred macaque is a cynomolgus monkey (Macaca fascicularis). Macaques (rhesus monkeys) are also contemplated.

In one aspect of the invention, the first binding domain binds to human EGFRVIII and further binds to macaque EGFRVIII, such as cynomolgus monkey EGFRVIII, and more preferably binds to macaque EGFRVIII expressed on the surface of a macaque cell. Preferred cynomolgus monkey EGFRVIII is shown in SEQ ID NO: 234. The affinity of the first binding domain for the cynomolgus EGFRVIII is preferably ≦ 15nM, more preferably ≦ 10nM, even more preferably ≦ 5nM, even more preferably ≦ 1nM, even more preferably ≦ 0.5nM, even more preferably ≦ 0.1nM, and most preferably ≦ 0.05nM or even ≦ 0.01 nM.

Preferably, the antibody construct according to the invention binds to the difference in affinity of macaque EGFRVIII relative to human EGFRVIII [ ma EGFRVIII: hu EGFRVIII ] (e.g. determined by BiaCore or by Scatchard analysis) is < 100, preferably < 20, more preferably < 15, further preferably < 10, even more preferably < 8, more preferably < 6, and most preferably < 2. Preferred ranges for the affinity gap for binding of the antibody construct according to the invention to cynomolgus monkey EGFRVIII relative to binding to human EGFRVIII are between 0.1 and 20, more preferably between 0.2 and 10, even more preferably between 0.3 and 6, even more preferably between 0.5 and 3 or between 0.5 and 2.5, and most preferably between 0.5 and 2 or between 0.6 and 2.

In one embodiment of the antibody construct of the invention, the second binding domain binds to human CD3 epsilon and to common marmoset, tamarisk villous or saimiri CD3 epsilon. Preferably, the second binding domain binds to an extracellular epitope of these CD3 epsilon chains. It is also envisaged that the second binding domain binds to an extracellular epitope of the human and cynomolgus CD3 epsilon chains. The most preferred epitope for CD3 epsilon is contained within amino acid residues 1-27 of the extracellular domain of human CD3 epsilon. Even more specifically, the epitope comprises at least the amino acid sequence Gln-Asp-Gly-Asn-Glu. Common marmosets and tamarix villosum are new world primates belonging to the family marmoset monkeys, whereas squirrel monkeys are new world primates belonging to the family urodeleidae.

It is particularly preferred for the antibody construct of the invention that the second binding domain that binds human CD3 on the surface of T cells comprises a VL region comprising a CDR-L1, CDR-L2 and CDR-L3 selected from the group consisting of:

(a) as shown in WO 2008/119567 SEQ ID NO: CDR-L1 shown in 27, SEQ ID NO as WO 2008/119567: 28 and the CDR-L2 as shown in WO 2008/119567 SEQ ID NO: CDR-L3 shown in FIG. 29;

(b) as shown in WO 2008/119567 SEQ ID NO: 117, CDR-L1 as set forth in SEQ ID NO of WO 2008/119567: 118 and the CDR-L2 as shown in WO 2008/119567 in SEQ ID NO: 119, CDR-L3; and

(c) as shown in WO 2008/119567 SEQ ID NO: 153, as set forth in SEQ ID NO: 154 and SEQ ID NO as shown in WO 2008/119567: 155, CDR-L3 shown.

In another preferred embodiment of the antibody construct of the invention, the second binding domain that binds human CD3 on the surface of a T cell comprises a VH region comprising a CDR-H1, CDR-H2 and CDR-H3 selected from the group consisting of:

(a) as shown in WO 2008/119567 SEQ ID NO:12, a CDR-H1 as set forth in SEQ ID NO of WO 2008/119567: 13 and the CDR-H2 as shown in WO 2008/119567 SEQ ID NO: CDR-H3 shown in FIG. 14;

(b) as shown in WO 2008/119567 SEQ ID NO:30, a CDR-H1 as set forth in SEQ ID NO of WO 2008/119567: 31 and the CDR-H2 as shown in WO 2008/119567 and SEQ ID NO: CDR-H3 shown in FIG. 32;

(c) as shown in WO 2008/119567 SEQ ID NO:48, a CDR-H1 as set forth in SEQ ID NO of WO 2008/119567: 49 and the CDR-H2 as shown in SEQ ID NO of WO 2008/119567: CDR-H3 shown in FIG. 50;

(d) as shown in WO 2008/119567 SEQ ID NO:66, a CDR-H1 as set forth in SEQ ID NO of WO 2008/119567: 67 and the CDR-H2 as set forth in SEQ ID NO of WO 2008/119567: 68 CDR-H3;

(e) as shown in WO 2008/119567 SEQ ID NO:84, a CDR-H1 as set forth in SEQ ID NO of WO 2008/119567: 85 and the CDR-H2 as shown in WO 2008/119567 as SEQ ID NO: CDR-H3 shown in 86;

(f) as shown in WO 2008/119567 SEQ ID NO:102, a CDR-H1 as set forth in SEQ ID NO of WO 2008/119567: 103 and a CDR-H2 as set forth in SEQ ID NO of WO 2008/119567: 104, CDR-H3;

(g) as shown in WO 2008/119567 SEQ ID NO: 120, a CDR-H1 as set forth in WO 2O08/119567 SEQ ID NO: 121 and the CDR-H2 as shown in WO 2008/119567 SEQ ID NO: 122, CDR-H3;

(h) as shown in WO 2008/119567 SEQ ID NO: 138, CDR-H1 as shown in WO 2008/119567 SEQ ID NO:139 and the CDR-H2 as shown in WO 2008/119567 SEQ ID NO:140, CDR-H3;

(i) as shown in WO 2008/119567 SEQ ID NO: 156, SEQ ID NO:157 and the CDR-H2 as shown in SEQ ID NO of WO 2008/119567: 158, CDR-H3; and

(j) as shown in WO 2008/119567 SEQ ID NO: 174, SEQ ID NO: 175 and SEQ ID NO as shown in WO 2008/119567 and CDR-H2 as shown in WO 2008/119567: 176, and CDR-H3.

It is further preferred for the antibody construct of the invention that the second binding domain that binds human CD3 on the surface of a T cell comprises a VL region selected from the group consisting of: as shown in SEQ ID NO: 18. SEQ ID NO: 27. SEQ ID NO: 36. SEQ ID NO: 45. SEQ ID NO: 54. SEQ ID NO: 63. SEQ ID NO: 72. SEQ ID NO: 81. SEQ ID NO: 90. SEQ ID NO:99 and SEQ ID NO:102 (see also SEQ ID NO:35, 39, 125, 129, 161 or 165 of WO 2008/119567).

It is further preferred that the second binding domain that binds human CD3 on the surface of a T cell comprises a VH region selected from the group consisting of: as shown in SEQ ID NO: 17. SEQ ID NO: 26. SEQ ID NO: 35. SEQ ID NO: 44. SEQ ID NO: 53. SEQ ID NO: 62. the amino acid sequence of SEQ ID NO: 71. SEQ ID NO: 80. SEQ ID NO: 89. SEQ ID NO:98 and SEQ ID NO:101 (see also SEQ ID NOs: 15, 19, 33, 37, 51, 55, 69, 73, 87, 91, 105, 109, 123, 127, 141, 145, 159, 163, 177 or 181 of WO 2008/119567).

More preferably, the antibody construct of the invention is characterized in that: the second binding domain that binds human CD3 on the surface of a T cell comprises a VL region and a VH region selected from the group consisting of:

(a) as shown in WO 2008/119567 SEQ ID NO:17 or 21 and the VL region as set forth in SEQ ID NO:15 or 19;

(b) as shown in WO 2008/119567 SEQ ID NO:35 or 39 and the VL region as set forth in SEQ ID NO:33 or 37;

(c) as shown in WO 2008/119567 SEQ ID NO:53 or 57 and the VL region as set forth in SEQ ID NO:51 or 55;

(d) as shown in WO 2008/119567 SEQ ID NO:71 or 75 and the VL region as set forth in SEQ ID NO:69 or 73;

(e) as shown in WO 2008/119567 SEQ ID NO:89 or 93 and the VL region as set forth in SEQ ID NO of WO 2008/119567: 87 or 91;

(f) as shown in WO 2008/119567 SEQ ID NO: 107 or 111 and the VL region as set forth in SEQ ID NO: 105 or 109; (g) as shown in WO 2008/119567 SEQ ID NO: 125 or 129 and a VL region as set forth in SEQ ID NO of WO 2008/119567: 123 or 127; (h) as shown in WO 2008/119567 SEQ ID NO:143 or 147 and the VL region as set forth in SEQ ID NO of WO 2008/119567: 141 or 145; (i) as shown in WO 2008/119567 SEQ ID NO: 161 or 165 and the VL region as set forth in SEQ ID NO:159 or 163;

(j) as shown in WO 2008/119567 SEQ ID NO: 179 or 183 and the VL region as set forth in SEQ ID NO of WO 2008/119567: the VH region shown in 177 or 181.

It is also preferred for the antibody construct of the invention that the second binding domain that binds human CD3 on the surface of a T cell comprises an amino acid sequence as set forth in SEQ ID NO:102 and a VL region as set forth in SEQ ID NO:101, VH region shown in figure 101.

According to a preferred embodiment of the antibody construct of the invention, the binding domain, in particular the second binding domain (which binds to human CD3 on the surface of T cells) has the following form: the paired VH and VL regions are in the form of single chain antibodies (scFv). The VH and VL regions are arranged in the order VH-VL or VL-VH. Preferably, the VH domain is located at the N-terminus of the linker sequence and the VL domain is located at the C-terminus of the linker sequence.

A preferred embodiment of the above antibody construct of the invention is characterized in that the second binding domain that binds to human CD3 on the surface of a T cell comprises an amino acid sequence selected from the group consisting of: SEQ ID NO: 19. the amino acid sequence of SEQ ID NO: 28. SEQ ID NO: 37. SEQ ID NO: 46. SEQ ID NO: 55. SEQ ID NO: 64. SEQ ID NO: 73. the amino acid sequence of SEQ ID NO: 82. SEQ ID NO: 91. SEQ ID NO:100 and SEQ ID NO:103 (see also SEQ ID NOs 23, 25, 41, 43, 59, 61, 77, 79, 95, 97, 113, 115, 131, 133, 149, 151, 167, 169, 185 or 187 of WO 2008/119567).

It is also envisaged that the antibody construct of the invention will have further functions in addition to its function of binding the target molecules EGFRVIII and CD 3. In this form, the antibody construct is a trifunctional or multifunctional antibody construct that functions by: targeting target cells via binding to EGFRVIII, modulating cytotoxic T cell activity via CD3 binding, and providing another function (such as a fully functional Fc constant domain that modulates antibody-dependent cellular cytotoxicity) via recruitment of effector cells such as NK cells, markers (fluorescent, etc.), therapeutic agents such as toxins or radionuclides, and/or substances that extend serum half-life, etc.

Examples of means to extend the serum half-life of the antibody constructs of the invention include peptides, proteins or protein domains fused or otherwise linked to the antibody construct. The group of peptides, proteins or protein domains includes peptides that bind to other proteins in the human body with preferred pharmacokinetic properties, such as serum albumin (see WO 2009/127691). Alternative concepts for such half-life extending peptides include peptides that bind to the neonatal Fc receptor (FcRn, see WO 2007/098420), which may also be used in the constructs of the invention. The concept of linking larger domains of proteins or complete proteins includes, for example, fusion of human serum albumin, variants or mutants of human serum albumin (see WO 2011/051489, WO 2012/059486, WO 2012/150319, WO 2013/135896, WO 2014/072481, WO 2013/075066) or domains thereof and fusion of immunoglobulin constant regions (Fc domains) and variants thereof. Such variants of Fc domains can be optimized/modified to allow for desired pairing of dimers or multimers, to eliminate Fc receptor binding (e.g., fey receptors) or for other reasons. Another concept known in the art to extend the half-life of small proteinaceous compounds in humans is the pegylation of compounds such as the antibody constructs of the invention.

In a preferred embodiment, the antibody construct of the invention is described as follows:

(a) a polypeptide comprising, in order from the N-terminus:

● has the sequence of SEQ ID NO: 159;

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 1-9; and

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 19. the amino acid sequence of SEQ ID NO: 28. SEQ ID NO: 37. SEQ ID NO: 46. SEQ ID NO: 55. SEQ ID NO: 64. SEQ ID NO: 73. SEQ ID NO: 82. SEQ ID NO: 91. SEQ ID NO:100 and SEQ ID NO: 103; and

● an optional His-tag, such as the His-tag shown in SEQ ID NO 10;

(b) a polypeptide comprising, in the following order from the N-terminus:

● has the sequence of SEQ ID NO: 159;

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 1-9;

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 19. SEQ ID NO: 28. SEQ ID NO: 37. SEQ ID NO: 46. SEQ ID NO: 55. SEQ ID NO: 64. SEQ ID NO: 73. the amino acid sequence of SEQ ID NO: 82. SEQ ID NO: 91. SEQ ID NO:100 and SEQ ID NO: 103;

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 104-134; and

● an optional His-tag, such as the His-tag shown in SEQ ID NO 10;

(c) a polypeptide comprising, in order from the N-terminus:

● polypeptide having the amino acid sequence QRFVTGHFGGLX1PANG (SEQ ID NO:135) wherein X1 is Y or H; and

● has the sequence of SEQ ID NO: 159;

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 19. SEQ ID NO: 28. SEQ ID NO: 37. the amino acid sequence of SEQ ID NO: 46. SEQ ID NO: 55. SEQ ID NO: 64. the amino acid sequence of SEQ ID NO: 73. the amino acid sequence of SEQ ID NO: 82. SEQ ID NO: 91. SEQ ID NO:100 and SEQ ID NO: 103;

● has an amino acid sequence of QRFVTGHFGGLHPANG (SEQ ID NO:137) or QRFCTGHFGGLHPCNG (SEQ ID NO: 139); and

an optional His-tag, such as the His-tag shown in SEQ ID NO 10;

(d) a polypeptide comprising, in the following order from the N-terminus:

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 17. SEQ ID NO: 26. SEQ ID NO: 35. SEQ ID NO: 44. SEQ ID NO: 53. SEQ ID NO: 62. SEQ ID NO: 71. SEQ ID NO: 80. SEQ ID NO: 89. SEQ ID NO:98 and SEQ ID NO: 101;

● has the sequence of SEQ ID NO: 8;

● has the sequence of SEQ ID NO: 158;

● has the sequence of SEQ ID NO: 140; and

a polypeptide comprising, in the following order from the N-terminus:

● has the sequence of SEQ ID NO: 157;

● has the sequence of SEQ ID NO: 8;

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 18. SEQ ID NO: 27. SEQ ID NO: 36. SEQ ID NO: 45. SEQ ID NO: 54. the amino acid sequence of SEQ ID NO: 63. SEQ ID NO: 72. SEQ ID NO: 81. SEQ ID NO: 90. SEQ ID NO:99 and SEQ ID NO:102 and a serine residue at the C-terminus;

● has the sequence of SEQ ID NO: 141;

(e) a polypeptide comprising, in order from the N-terminus:

● has the sequence of SEQ ID NO: 8;

● has the sequence of SEQ ID NO: 158;

● has the sequence of SEQ ID NO: 142; and

a polypeptide comprising, in order from the N-terminus:

● has the sequence of SEQ ID NO: 157;

● has the sequence of SEQ ID NO: 8;

has a sequence selected from the group consisting of SEQ ID NO: 18. SEQ ID NO: 27. SEQ ID NO: 36. SEQ ID NO: 45. SEQ ID NO: 54. SEQ ID NO: 63. SEQ ID NO: 72. SEQ ID NO: 81. SEQ ID NO: 90. SEQ ID NO:99 and SEQ ID NO:102 and a serine residue at the C-terminus;

● has the sequence of SEQ ID NO: 143;

(f) a polypeptide comprising, in order from the N-terminus:

● has the sequence of SEQ ID NO: 159;

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 19. SEQ ID NO: 28. SEQ ID NO: 37. SEQ ID NO: 46. SEQ ID NO: 55. SEQ ID NO: 64. SEQ ID NO: 73. SEQ ID NO: 82. SEQ ID NO: 91. SEQ ID NO:100 and SEQ ID NO: 103; and

● has the sequence of SEQ ID NO: 144; and a polypeptide having the sequence of SEQ ID NO: 145;

(g) a polypeptide comprising, in order from the N-terminus:

● has the sequence of SEQ ID NO: 159; and

● has the sequence of SEQ ID NO: 146; and

a polypeptide comprising, in order from the N-terminus:

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 19. SEQ ID NO: 28. SEQ ID NO: 37. SEQ ID NO: 46. SEQ ID NO: 55. the amino acid sequence of SEQ ID NO: 64. SEQ ID NO: 73. SEQ ID NO: 82. SEQ ID NO: 91. SEQ ID NO:100 and SEQ ID NO: 103; and

● has the sequence of SEQ ID NO: 147;

(h) a polypeptide comprising, in order from the N-terminus:

● has the sequence of SEQ ID NO: 159;

● has the sequence of SEQ ID NO: 148; and

a polypeptide comprising, in order from the N-terminus:

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 19. SEQ ID NO: 28. SEQ ID NO: 37. SEQ ID NO: 46. SEQ ID NO: 55. SEQ ID NO: 64. SEQ ID NO: 73. SEQ ID NO: 82. the amino acid sequence of SEQ ID NO: 91. SEQ ID NO:100 and SEQ ID NO: 103; and

● has the sequence of SEQ ID NO:149, or a polypeptide having an amino acid sequence set forth in seq id no; or

(i) A polypeptide comprising, in order from the N-terminus:

● has the sequence of SEQ ID NO: 159;

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 19. SEQ ID NO: 28. SEQ ID NO: 37. SEQ ID NO: 46. SEQ ID NO: 55. SEQ ID NO: 64. the amino acid sequence of SEQ ID NO: 73. SEQ ID NO: 82. SEQ ID NO: 91. SEQ ID NO:100 and SEQ ID NO: 103; and

● has the sequence of SEQ ID NO: 150;

(j) a polypeptide comprising, in order from the N-terminus:

● has the sequence of SEQ ID NO: 159;

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 19. SEQ ID NO: 28. SEQ ID NO: 37. SEQ ID NO: 46. SEQ ID NO: 55. SEQ ID NO: 64. SEQ ID NO: 73. SEQ ID NO: 82. SEQ ID NO: 91. SEQ ID NO:100 and SEQ ID NO: 103;

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 1.2, 4,5, 6,8 and 9; and

has a sequence selected from the group consisting of SEQ ID NO: 181-188.

As mentioned above, several preferred antibody constructs of the invention are modified by fusion with another moiety, such as albumin or an albumin variant. If these fusion constructs are characterized for their properties, in particular target affinity or cytotoxic activity, the skilled person will appreciate that similar fusion constructs or unmodified bispecific antibody constructs may be expected to have similar (or even better) properties. For example, if a bispecific antibody construct fused to albumin has an appreciable or desired cytotoxic activity or target affinity, it can be expected that the same/similar or even higher cytotoxic activity/target affinity will be observed for the construct without albumin.

According to another preferred embodiment, the bispecific antibody construct of the invention comprises (in addition to the two binding domains) a third domain comprising two polypeptide monomers, each comprising a hinge, a CH2 and a CH3 domain, wherein the two polypeptides (or polypeptide monomers) are fused to each other via a peptide linker. Preferably, the third domain comprises, in order from N-terminus to C-terminus: hinge-CH 2-CH 3-linker-hinge-CH 2-CH 3. Preferred amino acid sequences of the third domain are shown in SEQ ID NO: 181-188. Each of the two polypeptide monomers preferably has an amino acid sequence selected from the group consisting of SEQ ID NOs: 173-180 or an amino acid sequence which has at least 90% identity to those sequences. In another preferred embodiment, the first and second binding domains of the bispecific antibody construct of the invention are bound to the same epitope via a binding domain selected from the group consisting of SEQ ID NO: 1.2, 3,4, 5,6, 7, 8 and 9 is fused to the third domain.

According to the invention, a "hinge" is an IgG hinge region. This region can be identified by analogy using Kabat numbering, see Kabat position 223-. According to the above, the minimum requirement for a "hinge" is an IgG corresponding to D231 to P243 according to Kabat numbering₁The corresponding amino acid residues of the sequence segment. The terms CH2 and CH3 refer to immunoglobulin heavy chain constant regions 2 and 3. These regions can also be identified by analogy using Kabat numbering, see Kabat position 244-. It is understood that for its IgG₁Fc region, IgG₂Fc region, IgG₃Fc region, IgG₄There are some variations between immunoglobulins for the Fc region, IgM Fc region, IgA Fc region, IgD Fc region and IgE Fc region (see, e.g., Padlan, Molecular Immunology, 31(3), 169-217 (1993)). The term Fc monomer refers to the last two heavy chain constant regions of IgA, IgD, and IgG, and the last three heavy chain constant regions of IgE and IgM. The Fc monomer may also beIncluding flexible hinges at the N-terminus of these domains. For IgA and IgM, the Fc monomers may include J chains. For IgG, the Fc portion comprises the immunoglobulin domains CH2 and CH3 and the hinge between the first two domains and CH 2. Although the boundaries of the Fc portion of an immunoglobulin may vary, examples of human IgG heavy chain Fc portions comprising functional hinge, CH2, and CH3 domains may be defined, e.g., as comprising for IgG respectively₄D231 to (of the hinge domain) P476 (of the C-terminus of CH3 domain) or D231 to L476, where the numbering is according to Kabat.

Thus, the antibody construct of the invention may comprise, in order from N-terminus to C-terminus:

(a) a first binding domain;

(b) has a sequence selected from the group consisting of SEQ ID NO: SEQ ID NO: 1-9;

(c) a second binding domain;

(d) has a sequence selected from the group consisting of SEQ ID NO: 1.2, 4,5, 6,8 and 9;

(e) a first polypeptide monomer of a third domain (comprising the hinge, CH2, and CH3 domains);

(f) has a sequence selected from the group consisting of SEQ ID NO: 192. 193, 194, and 195; and

(g) a second polypeptide monomer of a third domain (comprising the hinge, CH2, and CH3 domains).

It is also preferred that the antibody construct of the invention comprises, in order from N-terminus to C-terminus:

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 159;

● having an amino acid sequence selected from the group consisting of: SEQ ID NO: 19. SEQ ID NO: 28. SEQ ID NO: 37. SEQ ID NO: 46. SEQ ID NO: 55. SEQ ID NO: 64. SEQ ID NO: 73. SEQ ID NO: 82. SEQ ID NO: 91. SEQ ID NO:100 and SEQ ID NO:103 (see also SEQ ID NOs 23, 25, 41, 43, 59, 61, 77, 79, 95, 97, 113, 115, 131, 133, 149, 151, 167, 169, 185 or 187 of WO 2008/119567);

● has an amino acid sequence selected from the group consisting of SEQ ID NO: 181-188.

Thus, in a preferred embodiment, the antibody construct of the invention comprises the amino acid sequence of SEQ ID NO:189 or 190 or consists of said polypeptide.

In one embodiment of the antibody construct of the invention, the antibody construct comprises the amino acid sequence as set forth in SEQ ID NO:160 or consists of said polypeptide.

Covalent modification of antibody constructs is also included within the scope of the invention and is typically, but not necessarily, performed post-translationally. For example, several types of covalent modifications of antibody constructs are introduced into the molecule by reacting specific amino acid residues of the antibody construct with an organic derivatizing agent capable of reacting with selected side chains or N-or C-terminal residues.

The cysteinyl residue is most often reacted with an α -haloacetate (and corresponding amine) such as chloroacetic acid or chloroacetamide to give a carboxymethyl or carboxamidomethyl derivative. Cysteinyl residues are also derivatized by reaction with: bromotrifluoroacetone, α -bromo- β - (5-imidazolyl) propionic acid, chloroacetyl phosphate, N-alkylmaleimide, 3-nitro-2-pyridyl disulfide, methyl 2-pyridyl disulfide, p-chloromercuribenzoate, 2-chloromercuriyl-4-nitrophenol, or chloro-7-nitrobenzo-2-oxa-1, 3-diazole.

Histidyl residues are derivatized by reaction with diethylpyrocarbonate at pH 5.5-7.0, since this agent is relatively specific for histidyl side chains. Para-bromobenzoylmethyl bromide is also useful; the reaction is preferably carried out in 0.1M sodium cacodylate at pH 6.0. Lysyl and amino terminal residues are reacted with succinic anhydride or other carboxylic acid anhydrides. Derivatization with these agents has the effect of reversing the charge of the lysyl residue. Other reagents suitable for derivatizing the alpha-amino containing residues include imidates such as picolineimine; pyridoxal phosphate; pyridoxal; a chloroborohydride compound; trinitrobenzenesulfonic acid; o-methylisourea; 2, 4-pentanedione; and transaminase-catalyzed reactions with glyoxylate.

Arginyl residues are modified by reaction with one or more conventional reagents, especially phenylglyoxal, 2, 3-butanedione, 1, 2-cyclohexanedione, and ninhydrin. Derivatization of arginine residues requires that the reaction be performed under basic conditions due to the high pKa of the guanidine functional group. In addition, these reagents can react with lysine groups as well as arginine epsilon-amino groups.

Specific modifications of tyrosyl residues can be performed, of particular interest is the introduction of spectroscopic tags into tyrosyl residues by reaction with aromatic diazo compounds or tetranitromethane. Most commonly, N-acetylimidazole and tetranitromethane are used to form O-acetyltyrosyl species and 3-nitro derivatives, respectively. By using¹²⁵I or¹³¹I iodination of tyrosyl residues to prepare labeled proteins for radioimmunoassay, the chloramine-T method described above is suitable.

Pendant carboxyl groups (aspartyl or glutamyl) are selectively modified by reaction with a carbodiimide (R ' -N ═ C ═ N — R '), where R and R ' are optionally different alkyl groups, for example 1-cyclohexyl-3- (2-morpholinyl-4-ethyl) carbodiimide or 1-ethyl-3- (4-azonia-4, 4-dimethylpentyl) carbodiimide. In addition, aspartyl and glutamyl residues are converted to asparaginyl and glutaminyl residues by reaction with ammonium ions.

Derivatization with bifunctional reagents can be used to crosslink the antibody constructs of the invention onto water-insoluble support matrices or surfaces for use in various methods. Commonly used cross-linking agents include, for example, 1-bis (diazoacetyl) -2-phenylethane, glutaraldehyde, N-hydroxysuccinimide esters (e.g., with 4-azidosalicylic acid), homobifunctional imide esters (including disuccinimidyl esters such as 3, 3' -dithiobis (succinimidyl propionate)), and bifunctional maleimides such as bis-N-maleimide-1, 8-octane. Derivatizing agents such as 3- [ (p-azidophenyl) dithio ] propinimomethyl ester produce photoactivated intermediates capable of forming crosslinks in the presence of light. Alternatively, protein immobilization is performed using a reactive water-insoluble matrix (e.g., cyanogen bromide activated carbohydrates) and a reactive substrate as described in U.S. Pat. Nos. 3,969,287, 3,691,016, 4,195,128, 4,247,642, 4,229,537, and 4,330,440.

Glutaminyl and asparaginyl residues are typically deamidated to the corresponding glutamyl and aspartyl residues, respectively. Alternatively, these residues are deamidated under mildly acidic conditions. Any form of these residues is within the scope of the present invention.

Other modifications include: hydroxylation of proline and lysine; phosphorylation of the hydroxyl group of seryl or threonyl residues; methylation of the alpha-amino group of lysine, arginine and histidine side chains (T.E.Creighton, Proteins: Structure and Molecular Properties, W.H.Freeman & Co., San Francisco, 1983, pp.79-86); acetylation of the N-terminal amine and amidation of any C-terminal carboxyl group.

Another type of covalent modification of antibody constructs included within the scope of the present invention includes altering the glycosylation pattern of the protein. As known in the art, the glycosylation pattern can depend on the sequence of the protein (e.g., the presence or absence of particular glycosylated amino acid residues discussed below) or the host cell or organism in which the protein is produced. Specific expression systems are discussed below.

Glycosylation of polypeptides is usually either N-linked or O-linked. N-linked refers to the attachment of a carbohydrate moiety to the side chain of an asparagine residue. The tripeptide sequences asparagine-X-serine and asparagine-X-threonine (where X is any amino acid except proline) are recognition sequences for enzymatic attachment of a carbohydrate moiety to an asparagine side chain. Thus, the presence of any of these tripeptide sequences in a polypeptide creates potential glycosylation sites. O-linked glycosylation refers to the attachment of one of the sugars N-acetylgalactosamine, galactose or xylose to a hydroxyamino acid, with serine or threonine being most commonly used, although 5-hydroxyproline or 5-hydroxylysine may also be used.

The addition of glycosylation sites (directed to N-linked glycosylation sites) to the antibody construct is conveniently accomplished by altering the amino acid sequence so that it contains one or more of the above-described tripeptide sequences. Changes (to O-linked glycosylation sites) can also be made by adding or replacing one or more serine or threonine residues to the starting sequence. For convenience, the amino acid sequence of the antibody construct is preferably altered by alteration at the DNA level, particularly by mutating the DNA encoding the polypeptide at preselected bases, to create codons that will be translated into the desired amino acids.

Another means of increasing the number of carbohydrate moieties on an antibody construct is by chemical or enzymatic coupling of glycosides to the protein. These methods are advantageous because they do not require the production of proteins in host cells that have glycosylation capacity for N-linked and O-linked glycosylation. Depending on the coupling scheme used, the sugar may be attached to (a) arginine and histidine; (b) a free carboxyl group; (c) free sulfhydryl groups such as cysteine sulfhydryl groups; (d) a free hydroxyl group such as that of serine, threonine or hydroxyproline; (e) aromatic residues such as phenylalanine, tyrosine or tryptophan; or (f) the amide group of glutamine. These methods are described in WO 87/05330 and Aplin and Wriston, 1981, CRC Crit. Rev. biochem., p.259-306.

Removal of the carbohydrate moiety present on the starting antibody construct may be accomplished by chemical or enzymatic means. Chemical deglycosylation requires exposing the protein to the compound triflic acid or equivalent compound. This treatment results in the cleavage of most or all of the sugars other than the linked sugar (N-acetylglucosamine or N-acetylgalactosamine) while leaving the polypeptide intact. Hakimuddin et al, 1987, arch, biochem, biophysis, 259: 52 and Edge et al, 1981, anal. biochem.118: 131 describes chemical deglycosylation. Enzymatic cleavage of the carbohydrate moiety on a polypeptide can be achieved by using various endoglycosidases and exoglycosidases, such as Thotakura et al, 1987, meth.enzymol.138: 350. As described by Duskin et al, 1982, J.biol.chem.257: 3105, glycosylation of potential glycosylation sites can be prevented by using tunicamycin compounds. Tunicamycin blocks the formation of protein-N-glycosidic bonds.

Other modifications of the antibody constructs are also contemplated herein. For example, another type of covalent modification of the antibody construct includes linking the antibody construct to various non-protein polymers, including but not limited to various polyols such as polyethylene glycol, polypropylene glycol, polyalkylene oxide, or copolymers of polyethylene glycol and polypropylene glycol, in the manner described in U.S. Pat. nos. 4,640,835, 4,496,689, 4,301,144, 4,670,417, 4,791,192, or 4,179,337. In addition, amino acid substitutions may be made at various positions within the antibody construct, for example to facilitate addition of a polymer such as PEG, as is known in the art.

In some embodiments, the covalent modification of the antibody construct of the invention comprises the addition of one or more labels. The labeling group may be coupled to the antibody construct via spacer arms of various lengths, thereby reducing potential steric hindrance. Various methods for labeling proteins are known in the art and can be used in the practice of the present invention. The term "label" or "labeling group" refers to any detectable label. Generally, the labels fall into a variety of categories depending on the assay in which the label is detected, examples include, but are not limited to:

a) an isotopic label, which can be a radioisotope or a heavy isotope, such as a radioisotope or radionuclide (e.g.,³H、¹⁴C、¹⁵N、³⁵S、⁸⁹Zr、⁹⁰Y、⁹⁹Tc、¹¹¹In、¹²⁵I、¹³¹I)

b) magnetic labels (e.g. magnetic particles)

c) Redox active moieties

d) Optical dyes (including but not limited to chromophores, phosphors, and fluorophores), such as fluorophores (e.g., FITC, rhodamine, lanthanide phosphors), chemiluminescent groups, and fluorophores that can be "small molecule" fluorescein or protein fluorescein

e) Enzymatic groups (e.g., horseradish peroxidase,. beta. -galactosidase, luciferase, alkaline phosphatase)

f) Biotinylation group

g) Predetermined polypeptide epitopes recognized by secondary reporters (e.g., leucine zipper pair sequences, binding sites for secondary antibodies, metal binding domains, epitope tags, etc.)

"fluorescent label" refers to any molecule that can be detected by its inherent fluorescent properties. Suitable fluorescent labels include, but are not limited to, fluorescein, rhodamine, tetramethylrhodamine, eosin, erythrosine, coumarin, methylcoumarin, pyrene, malachite green (Malcite green), stilbene, lucifer yellow, cascade Blue J (Cascade Blue J), Texas Red (Texas Red), IAEDNS, EDANS, BODIPY FL, LC Red 640, Cy5, Cy5.5, LC Red 705, Oregon green (Oregon green), Alexa-Fluor dyes (Alexa Fluor 350, Alexa Fluor 430, Alexa Fluor 488, Alexa Fluor 546, Alexa Fluor 568, Alexa Fluor 594, Alexa Fluor 633, Alexa Fluor 660, Alexa Fluor 680), Cascade blue, Cascade Yellow (Cascade Yellow) and R-Phycoerythrin (PE) (Molecular Probes, Eugene, OR), FITC, rhodamine and texas Red (Pierce, Rockford, IL), Cy5, Cy5.5, Cy7(Amersham Life Science, Pittsburgh, PA). Suitable optical dyes (including fluorophores) are described in Molecular Probes Handbook, Richard p.

Suitable fluorescent labels for proteins also include, but are not limited to, green fluorescent proteins, including GFP, Renilla (Renilla), Penaeus (Ptilosarcus) or Aequorea (Aequorea) species (Chalfie et al, 1994, Science 263: 802-); EGFP (Clontech Laboratories, Inc, Genbank accession No. U55762); blue fluorescent protein (BFP, Quantum Biotechnologies, Inc.1801 de Maison neuve Blvd. West, 8th Floor, Montreal, Quebec, Canada H3H 1J 9; Stauber, 1998, Biotechnologies 24: 462-471; Heim et al, 1996, curr. biol. 6: 178-); enhanced yellow fluorescent protein (EYFP, Clontech Laboratories, Inc.); luciferase (Ichiki et al, 1993, J.Immunol.150: 5408-5417); beta-galactosidase (Nolan et al, 1988, Proc. Natl. Acad. Sci. U.S.A.85: 2603-No. 2607) and Renilla (WO92/15673, WO95/07463, WO98/14605, WO98/26277, WO 99/49019; U.S. Pat. No. 5,292,658; 5,418,155; 5,683,888; 5,741,668; 5,777,079; 5,804,387; 5,874,304; 5,876,995; 5,925,558).

Leucine zipper domain is a peptide that promotes oligomerization of proteins comprising the leucine zipper domain. Leucine zippers were originally identified in several DNA binding proteins (Landshulz et al, 1988, Science 240: 1759) and since then discovered in a variety of different proteins. Among the known leucine zippers are dimeric or trimeric naturally occurring peptides and derivatives thereof. Examples of leucine zipper domains suitable for use in generating soluble oligomeric proteins are described in PCT application WO 94/10308, and Hoppe et al, 1994, FEBS Letters 344: 191 describes a leucine zipper derived from lung surfactant protein d (spd). In Fanslow et al, 1994, semin. immunol.6: 267-78, to allow stable trimerization of heterologous proteins fused thereto. In one method, a recombinant fusion protein comprising an EGFRVIII antibody fragment or derivative fused to a leucine zipper peptide is expressed in a suitable host cell, and the resulting soluble oligomeric EGFRVIII antibody fragment or derivative is recovered from the culture supernatant.

The antibody constructs of the invention may also comprise additional domains which, for example, aid in the isolation of the molecule or are associated with an adaptive pharmacokinetic profile of the molecule. The domains that facilitate isolation of the antibody construct may be selected from peptide motifs or secondary import moieties that can be captured in an isolation procedure such as a separation column. Non-limiting embodiments of such additional domains include peptide motifs referred to as Myc tags, HAT tags, HA tags, TAP tags, GST tags, chitin binding domains (CBD tags), maltose binding proteins (MBP tags), Flag tags, Strep tags and variants thereof (e.g., StrepII tags), and His tags. All antibody constructs disclosed herein characterized by the identified CDRs preferably comprise a His-tag domain, which is generally referred to as a repeat sequence of consecutive His residues, preferably five and more preferably six His residues (hexa-histidine) in the amino acid sequence of the molecule. The His-tag may be located, for example, at the N-terminus or C-terminus of the antibody construct, which is preferably located at the C-terminus. Most preferably, the hexahistidine tag (HHHHHHHHHH) (SEQ ID NO: 10) is linked via a peptide bond to the C-terminus of the antibody construct according to the invention.

The first binding domain of the antibody construct of the invention binds to human EGFRVIII on the surface of a target cell. A preferred amino acid sequence of human EGFRVIII consists of NO: 231. 232 and 233. Thus, the first binding domain according to the invention preferably binds to EGFRVIII when the EGFRVIII is expressed by a naturally expressing cell or cell line and/or by a cell or cell line transformed or (stably/transiently) transfected with EGFRVIII. In preferred embodiments, when EGFRVIII is used as a "target" or "ligand" molecule in an in vitro binding assay such as BIAcore or Scatchard, the first binding domain also binds to EGFRVIII. A "target cell" can be any prokaryotic or eukaryotic cell expressing EGFRVIII on its surface; preferably, the target cell is a cell of a part of the human or animal body, such as ovarian cancer cells, pancreatic cancer cells, mesothelioma cells, lung cancer cells, gastric cancer cells and triple negative breast cancer cells.

The affinity of the first binding domain for human EGFRVIII is preferably ≦ 20nM, more preferably ≦ 10nM, even more preferably ≦ 5nM, even more preferably ≦ 2nM, even more preferably ≦ 1nM, even more preferably ≦ 0.6nM, even more preferably ≦ 0.5nM, and most preferably ≦ 0.4 nM. Affinity can be measured, for example, in a BIAcore assay or Scatchard assay, e.g., as described in the examples. Other methods of determining affinity are also well known to the skilled person; see, for example, the accompanying examples.

Amino acid sequence modifications of the antibody constructs described herein are also contemplated. For example, it may be desirable to improve the binding affinity and/or other biological properties of the antibody construct. Amino acid sequence variants of the antibody construct are prepared by introducing appropriate nucleotide changes into the antibody construct nucleic acid or by peptide synthesis. All amino acid sequence modifications described below should result in antibody constructs that still retain the desired biological activity of the unmodified parent molecule (binding to EGFRVIII and CD 3).

The term "amino acid" or "amino acid residue" generally refers to an amino acid having its art-recognized definition, e.g., an amino acid selected from the group consisting of: alanine (Ala or a); arginine (Arg or R); asparagine (Asn or N); aspartic acid (Asp or D); cysteine (Cys or C); glutamine (Gln or Q); glutamic acid (Glu or E); glycine (Gly or G); histidine (His or H); isoleucine (He or I); leucine (Leu or L); lysine (Lys or K); methionine (Met or M); phenylalanine (Phe or F); proline (Pro or P); serine (Ser or S); threonine (Thr or T); tryptophan (Trp or W); tyrosine (Tyr or Y); and valine (Val or V), but modified, synthetic or unusual amino acids may also be used as desired. Generally, amino acids can be grouped as having a non-polar side chain (e.g., Ala, Cys, He, Leu, Met, Phe, Pro, Val); negatively charged side chains (e.g., Asp, Glu); positively charged side chains (e.g., Arg, His, Lys); or uncharged polar side chains (e.g., Asn, Cys, Gln, Gly, His, Met, Phe, Ser, Thr, Trp, and Tyr).

Amino acid modifications include, for example, deletions from and/or insertions into and/or substitutions of residues within the amino acid sequence of the antibody construct. Any combination of deletions, insertions, and substitutions can be made to arrive at the final construct, so long as the final construct possesses the desired characteristics. Amino acid changes may also alter post-translational processes of the antibody construct, such as changing the number or position of glycosylation sites.

For example, 1, 2,3, 4,5, or 6 amino acids may be inserted or deleted in each CDR (depending, of course, on the length of the CDR), while 1, 2,3, 4,5, 6,7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or 25 amino acids may be inserted or deleted in each FR. Preferably, amino acid sequence insertions include amino and/or carboxy terminal fusions of 1, 2,3, 4,5, 6,7, 8, 9, or 10 residues in length, as well as intrasequence insertions of single or multiple amino acid residues, which are inserted into a polypeptide containing one hundred or more residues. Insertional variants of the antibody constructs of the invention include fusions to the N-or C-terminus of the antibody construct with an enzyme or to a polypeptide that extends the serum half-life of the antibody construct.

The sites of substitutional mutagenesis of most interest include the CDRs, particularly the hypervariable regions, of the heavy and/or light chain, but FR alterations of the heavy and/or light chain are also contemplated. Substitutions are preferably conservative substitutions as described herein. Preferably, 1, 2,3, 4,5, 6,7, 8, 9 or 10 amino acids may be substituted in the CDRs and 1, 2,3, 4,5, 6,7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or 25 amino acids may be substituted in the Framework Regions (FRs), depending on the length of the CDRs or FRs. For example, if a CDR sequence comprises 6 amino acids, substitutions of one, two or three of these amino acids are envisaged. Similarly, if the CDR sequence comprises 15 amino acids, substitutions of 1, 2,3, 4,5 or 6 of these amino acids are envisaged.

A useful method for identifying certain residues or regions of an antibody construct as preferred mutagenesis positions is referred to as "alanine scanning mutagenesis", e.g. as described by Cunningham and Wells, Science, 244: 1081 and 1085 (1989). Here, one residue or group of target residues within an antibody construct (e.g., charged residues such as arg, asp, his, lys, and glu) is identified and replaced with a neutral or negatively charged amino acid (most preferably alanine or polyalanine) to affect the interaction of the amino acid with the epitope.

Those amino acid positions exhibiting functional sensitivity to substitution are then precisely identified by introducing additional or other variants at or for the substitution site. Thus, while the site or region for introducing amino acid sequence changes is predetermined, the nature of the mutation itself need not be predetermined. For example, to analyze or optimize the performance of a mutation at a given site, alanine scanning or random mutagenesis can be performed at the target codon or region, and the expressed antibody construct variants screened for the optimal combination of desired activities. Techniques for making substitution mutations at predetermined sites in DNA having a known sequence are well known, such as M13 primer mutagenesis and PCR mutagenesis. Assays for antigen binding activity (e.g., EGFRVIII or CD3 binding) are used to screen mutants.

Generally, if amino acids in one or more or all CDRs of the heavy and/or light chain are substituted, it is preferred that the subsequently obtained "substituted" sequence has at least 60% or 65%, more preferably 70% or 75%, even more preferably 80% or 85%, and particularly preferably 90% or 95% identity with the "original" CDR sequence. This means that the degree of identity with the "substituted" sequence depends on the length of the CDR. For example, a CDR having 5 amino acids preferably has 80% identity to its replacement sequence in order to replace at least one amino acid. Thus, the CDRs of an antibody construct may have varying degrees of identity with their replacement sequences, for example CDRL1 may have 80% and CDRL3 may have 90%.

Preferred substitutions (or substitutions) are conservative substitutions. However, any substitution (including non-conservative substitutions or one or more from the "exemplary substitutions" listed in table 1 below) is contemplated as long as the antibody construct retains its ability to bind EGFRVIII via the first binding domain and CD3 or CD3 epsilon via the second binding domain and/or its CDRs are identical (at least 60% or 65%, more preferably 70% or 75%, even more preferably 80% or 85%, and particularly preferably 90% or 95% identical to the "original" CDR sequences) to the subsequent substitution sequences.

Conservative substitutions are shown in table 1 under the heading "preferred substitutions". If these substitutions result in a change in biological activity, more substantial changes, referred to as "exemplary substitutions" in Table 1 or as described further below with respect to amino acid classes, can be introduced and the products screened for desired characteristics.

Table 1: amino acid substitutions

Substantial changes in the biological properties of the antibody constructs of the invention can be achieved by selecting substitutions that differ significantly in their effect on maintaining (a) the structure, e.g., the folded or helical conformation, of the polypeptide backbone in the region of the substitution, (b) the charge or hydrophobicity of the molecule at the target site or (c) the volume of the side chain. Naturally occurring residues may be grouped according to common side chain properties: (1) hydrophobic residue: norleucine, met, ala, val, leu, ile; (2) neutral hydrophilic residues: cys, ser, thr; (3) acidic residue: asp, glu; (4) basic residue: asn, gin, his, lys, arg; (5) residues that influence chain orientation: gly, pro; and (6) aromatic residues: trp, tyr, phe.

Non-conservative substitutions will require the replacement of one of these classes for another. Any cysteine residues not involved in maintaining the appropriate conformation of the antibody construct may typically be replaced with serine to improve the oxidative stability of the molecule and prevent aberrant cross-linking. Conversely, cysteine bonds may be added to the antibody to improve its stability (particularly where the antibody is an antibody fragment such as an Fv fragment).

For amino acid sequences, sequence identity and/or similarity is determined by using standard techniques known in the art, including, but not limited to, the local sequence identity algorithm (Smith and Waterman, 1981, adv. appl. math.2: 482); sequence identity alignment algorithms (Needleman and Wunsch, 1970, J.mol.biol.48: 443); similar search methods (Pearson and Lipman, 1988, Proc. Nat. Acad. Sci. U.S.A.85: 2444); computer implementation of these algorithms (GAP, BESTFIT, FASTA and TFASTA in Wisconsin Genetics Software Package, Genetics Computer Group, 575 Science Drive, Madison, Wis.); the method is described by Devereux et al, 1984, nucleic acid res.12: the Best Fit sequence program described in 387- & 395 is preferably set using default settings or passed. Preferably, the percent identity is calculated by FastDB based on the following parameters: a mismatch penalty of 1; gap penalty 1; gap size penalty 0.33; and a connection penalty of 30 ("Current Methods in Sequence company and Analysis", macromolecular Sequencing and Synthesis, Selected Methods and Applications, pp.127-149 (1988), Alan R.Liss, Inc.

An example of an available algorithm is PILEUP. PILEUP creates multiple sequence alignments from a set of related sequences using progressive pairwise alignments. It can also plot a dendrogram showing the clustering relationships used to generate the alignment. PILEUP uses Feng and Doolittle, 1987, j.mol.evol.35: 351-360 simplified form of the progressive comparison method; the method is similar to Higgins and Sharp, 1989, cabaos 5: 151-153. Useful PILEUP parameters include a default gap weight of 3.00, a default gap length weight of 0.10, and weighted end gaps.

Another example of a useful algorithm is the BLAST algorithm, which is described in: altschul et al, 1990, j.mol.biol.215: 403-; altschul et al, 1997, Nucleic Acids Res.25: 3389 and 3402; and Karin et al, 1993, proc.natl.acad.sci.u.s.a.90: 5873 in 5787. Particularly useful BLAST programs are available from Altschul et al, 1996, Methods in Enzymology 266: 460-480 WU-BLAST-2 program. WU-BLAST-2 uses several search parameters, most of which are set to default values. The adjustable parameters were set using the following values: overlap span 1; overlap fraction is 0.125; the word threshold (T) ═ II. The HSP S and HSP S2 parameters are dynamic values and are determined by the program itself based on the composition of the particular sequence and the composition of the particular database in which the target sequence is searched; however, these values may be adjusted to improve sensitivity.

Another algorithm that can be used is the algorithm described by Altschul et al, 1993, nucleic acids res.25: 3389-3402 (gapped BLAST). Gap BLAST uses BLOSUM-62 substitution scoring; the threshold T parameter is set to 9; two-hit method (two-hit method) is used to trigger no-vacancy extension, charge vacancy length k, at the cost of 10+ k; xu is set to 16, and Xg is set to 40 for the database search phase and 67 for the algorithm output phase. Gap alignments are triggered by scores corresponding to about 22 bits.

Typically, the amino acid homology, similarity or identity between the individual variant CDRs and the sequences described herein is at least 60%, and more typically has an increasing homology or identity of preferably at least 65% or 70%, more preferably at least 75% or 80%, even more preferably at least 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% and almost 100%. Similarly, "percent (%) nucleic acid sequence identity" with respect to a nucleic acid sequence of a binding protein identified herein is defined as the percentage of nucleotide residues in the candidate sequence that are identical to the nucleotide residues in the coding sequence of the antibody construct. A specific approach is to use the BLASTN module of WU-BLAST-2 set to default parameters, where the overlap span and overlap score are set to 1 and 0.125, respectively.

Typically, the nucleotide sequence encoding each variant CDR has at least 60% nucleic acid sequence homology, similarity or identity with the nucleotide sequences described herein, and more typically has preferably at least 65%, 70%, 75%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% and almost 100% incremental homology or identity. Thus, a "variant CDR" is a variant having a specified homology, similarity or identity to a parent CDR of the invention and sharing a biological function, including but not limited to at least 60%, 65%, 70%, 75%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% of the specificity and/or activity of the parent CDR.

In one embodiment, the percentage identity of the antibody construct according to the invention to the human germline is ≥ 70% or ≥ 75%, more preferably ≥ 80% or ≥ 85%, even more preferably ≥ 90%, and most preferably ≥ 91%, ≥ 92%, ≥ 93%, ≥ 94%, ≥ 95% or even ≥ 96%. Identity with human antibody germline gene products is considered an important feature to reduce the risk of a therapeutic protein eliciting a patient's immune response to a drug during treatment. Hwang and Foote ("Immunogenicity of engineered antibodies"; Methods 36(2005)3-10) demonstrated that reducing the non-human portion of the drug antibody construct can reduce the risk of inducing anti-drug antibodies in patients during treatment. By comparing the exhaustive number of clinically evaluated antibody drugs with the corresponding immunogenicity data, the trend was found that humanization of the antibody V regions resulted in a protein with lower immunogenicity (on average 5.1% of patients) than antibodies carrying unaltered non-human V regions (on average 23.59% of patients). Thus, a higher degree of identity to human sequences is desirable for V-region based protein therapies in the form of antibody constructs. To determine germline identity, the V region of the VL can be aligned with the amino acid sequences of the human germline V and J segments (http:// vbase. mrc-cpe.cam.ac.uk /) using Vector NTI software, and the amino acid sequences calculated as a percentage by dividing the same amino acid residues by the total number of amino acid residues of the VL. The same is true for the VH segment (http:// vbase. mrc-cpe. cam. ac. uk /), although VH CDR3 can be excluded due to its high diversity and lack of existing human germline VH CDR3 alignment partners. Recombinant techniques can then be used to increase sequence identity to human antibody germline genes.

In another embodiment, the bispecific antibody constructs of the invention exhibit high monomer yields under standard research-grade conditions (e.g., in a standard two-step purification process). Preferably, the monomer yield of the antibody construct according to the invention is ≥ 0.25mg/L supernatant, more preferably ≥ 0.5mg/L supernatant, even more preferably ≥ 1mg/L supernatant, and most preferably ≥ 3mg/L supernatant.

Likewise, the yield of the dimeric antibody construct isotype, and thus the percentage of monomers of the antibody construct (i.e., monomers (monomer + dimer)), can be determined. Yield and percentage monomer calculations for both monomeric and dimeric antibody constructs can be obtained, for example, in SEC purification steps from culture supernatants from standardized research-grade production in roller bottles. In one embodiment, the monomer percentage of the antibody construct is 80% or more, more preferably 85% or more, even more preferably 90% or more, and most preferably 95% or more.

In one embodiment, the antibody construct has a preferred plasma stability (ratio of EC50 with plasma to EC50 without plasma) of ≦ 5 or ≦ 4, more preferably ≦ 3.5 or ≦ 3, even more preferably ≦ 2.5 or ≦ 2, and most preferably ≦ 1.5 or ≦ 1. The plasma stability of the antibody constructs can be tested by incubating the constructs in human plasma for 24 hours at 37 ℃ followed by an EC50 assay in a 51-chromium release cytotoxicity assay. The effector cells in the cytotoxicity assay can be stimulated enriched human CD8 positive T cells. The target cell may, for example, be a CHO cell transfected with human EGFRVIII. The ratio of effector cells to target cells (E: T) can be selected to be 10: 1. The human plasma pool used for this purpose was derived from blood of healthy donors collected by EDTA-coated syringes. The cellular components were removed by centrifugation and the upper plasma phase was collected and then pooled. As a control, the antibody construct was diluted in RPMI-1640 medium immediately prior to the cytotoxicity assay. Plasma stability was calculated as the ratio of EC50 (after plasma incubation) to EC50 (control). See example 7.

It is also preferred that the antibody constructs of the invention have a low monomer-dimer conversion. Conversion can be measured under different conditions and analyzed by high performance size exclusion chromatography. See example 5. For example, the monomeric isoform of the antibody construct may be incubated at 37 ℃ and a concentration of 250. mu.g/ml for 7 days in an incubator. Under these conditions, the antibody construct of the present invention preferably has a dimer percentage of ≦ 2.5%, more preferably ≦ 2%, further preferably ≦ 1.5%, further preferably ≦ 1%, more preferably ≦ 0.5% and most preferably ≦ 0.25%. Whereas the ev iii-2 based bispecific antibody construct showed a dimer conversion of 1.56%, the ev iii-1 based bispecific antibody construct of the present invention showed a conversion at the upper limit of the most preferred limit for this feature.

It is also preferred that the bispecific antibody constructs of the invention exhibit extremely low dimer conversion after multiple freeze-thaw cycles. For example, antibody construct monomers are adjusted to a concentration of 250 μ g/ml in, for example, a common formulation buffer, and subjected to three freeze-thaw cycles (freezing at-80 ℃ for 30 minutes followed by thawing at room temperature for 30 minutes), followed by high-efficiency SEC assays to determine the percentage of the original monomeric antibody construct that has been converted to a dimeric antibody construct. For example, after three freeze-thaw cycles, the percentage of dimer of the bispecific antibody construct is preferably ≦ 2.5%, more preferably ≦ 2%, further preferably ≦ 1.5%, further preferably ≦ 1%, and most preferably ≦ 0.5%. Whereas the ev iii-2 based bispecific antibody construct did not meet the preferred range (2.53% dimer), the ev iii-1 based bispecific antibody construct of the present invention showed a conversion rate (0.59% dimer) at the upper limit of the most preferred limit for this feature.

The bispecific antibody constructs of the present invention preferably show a favourable thermostability in which the aggregation temperature is ≥ 45 ℃ or ≥ 50 ℃, more preferably ≥ 51 ≥ 52 ≥ 53 ≥ 54 ℃, even more preferably ≥ 56 ℃ or ≥ 57 ℃ and most preferably ≥ 58 ℃ or ≥ 59 ℃. The thermostability parameter can be determined from the antibody aggregation temperature as follows: the antibody solution at a concentration of 250. mu.g/ml was transferred to a disposable cuvette and placed in a Dynamic Light Scattering (DLS) device. The sample was heated from 40 ℃ to 70 ℃ at a heating rate of 0.5 ℃/min and radius measurements were continuously taken. The increase in radius indicating melting and aggregation of the protein was used to calculate the aggregation temperature of the antibody. See example 6.

Alternatively, the temperature melting curve can be determined by Differential Scanning Calorimetry (DSC) to determine the intrinsic biophysical protein stability of the antibody construct. These experiments were performed using a MicroCal LLC (normampton, MA, u.s.a) VP-DSC apparatus. The energy absorption of the samples containing the antibody construct was recorded from 20 ℃ to 90 ℃ compared to the samples containing the formulation buffer alone. The antibody construct is adjusted to a final concentration of 250 μ g/ml in, for example, SEC running buffer. The overall sample temperature was gradually increased in order to record the corresponding melting curve. The energy absorption of the sample and the formulation buffer reference was recorded at each temperature T. The difference of the energy absorption Cp (kcal/mole/° c) of the sample minus the reference value is plotted against the corresponding temperature. The melting temperature is defined as the temperature at which the first maximum of energy absorption occurs.

It is also contemplated that the EGFRUIXICd 3 bispecific antibody construct of the invention has a turbidity (measured by OD340 after concentrating the purified monomeric antibody construct to 2.5mg/ml and incubating overnight) of ≦ 0.2, preferably ≦ 0.15, more preferably ≦ 0.12, even more preferably ≦ 0.1 or even ≧ 0.09, and most preferably ≦ 0.08 or ≧ 0.07. See example 7. The bispecific antibody construct EvIII-2 showed a rather high turbidity of almost 3 (measured by OD340 after concentration of the purified monomeric antibody construct to 2.5mg/ml and incubation overnight), whereas only the bispecific antibody construct based on EvIII-1 was clearly within the range of the desired proteinaceous compound useful for pharmaceutical formulations; see table 3 in example 7.

In another embodiment, the antibody construct according to the invention is stable at acidic pH. The higher the tolerance exhibited by the antibody construct at a non-physiological pH, such as pH5.5 (the pH required to perform, for example, cation exchange chromatography), the higher the recovery of the antibody construct eluted from the ion exchange column relative to the total amount of protein loaded. The recovery of the antibody construct from an ion (e.g., cation) exchange column at pH5.5 is preferably ≥ 30%, more preferably ≥ 40%, more preferably ≥ 50%, even more preferably ≥ 60%, even more preferably ≥ 70%, even more preferably ≥ 80%, even more preferably ≥ 90%, even more preferably ≥ 95%, and most preferably ≥ 99%.

Furthermore, it is envisaged that the bispecific antibody constructs of the present invention exhibit therapeutic efficacy or anti-tumor activity. This can be assessed, for example, in studies as disclosed in the following example of an advanced human tumor xenograft model:

on day 1 of the study, 5X 10 of human EGFRVIII-positive cancer cell lines (e.g., human glioblastoma cell line U87 or DK-MG)⁶Individual cells were injected subcutaneously into the right dorsal side of female NOD/SCID mice. When the mean tumor volume reached about 100mm³By mixing about 2X 10⁷Individual cells were injected into the abdominal cavity of animals, and human CD3 positive T cells expanded in vitro were transplanted into mice. The mice of vehicle control group 1 received no effector cells and served as non-transplant controls compared to vehicle control group 2 (receiving effector cells) to monitor the effect of T cells alone on tumor growth. When the mean tumor volume reached about 200mm³At that time, antibody therapy was initiated. The mean tumor size of each treatment group should not be statistically different from any other group on the day of treatment initiation (analysis of variance). Mice were treated with 0.5 mg/kg/day of the EGFRVIIIxCD3 bispecific antibody construct by bolus injection intravenously for about 15 to 20 days. Tumors were measured by calipers during the study and progression was assessed by inter-group comparison of Tumor Volume (TV). Determination of tumor growth inhibition T/C [% ] by calculating TV as T/C% ═ 100 × (median TV of assay group)/(median TV of control group 2)]。

The skilled person knows how to modify or adjust certain parameters of the present study, such as the number of tumor cells injected, the injection site, the number of human T cells transplanted, the amount of bispecific antibody construct to be administered, and the time axis, while still achieving meaningful and reproducible results. Preferably, the tumor growth inhibition T/C [% ] is 70 or 60 or less, more preferably 50 or 40 or less, even more preferably 30 or 20 or less, and most preferably 10 or 5 or even 2.5 or less.

The invention further provides polynucleotide/nucleic acid molecules encoding the antibody constructs of the invention.

Polynucleotides are biopolymers consisting of 13 or more nucleotide monomers covalently bound in a chain. DNA (e.g., cDNA) and RNA (e.g., mRNA) are examples of polynucleotides having different biological functions. Nucleotides are organic molecules that function as monomers or subunits of nucleic acid molecules such as DNA or RNA. The nucleic acid molecule or polynucleotide may be double-stranded and single-stranded, linear and circular. It is preferably contained in a vector, which is preferably contained in a host cell. The host cell is capable of expressing the antibody construct, for example, after transformation or transfection with a vector or polynucleotide of the invention. To this end, the polynucleotide or nucleic acid molecule is operably linked to a control sequence.

The genetic code is a set of rules whereby information encoded in genetic material (nucleic acids) is translated into protein. Biological decoding in living cells is achieved by ribosomes, which use tRNA molecules to carry amino acids and read three nucleotides of mRNA at a time to join the amino acids in the order specified by the mRNA. The code defines how the sequence of these nucleotide triplets (called codons) specifies which amino acid is to be added next during protein synthesis. With some exceptions, a trinucleotide codon in a nucleic acid sequence specifies a single amino acid. Since most genes are encoded with the exact same code, this particular code is often referred to as the canonical or standard genetic code. While the genetic code determines the protein sequence of a given coding region, other genomic regions can influence the time and place of production of these proteins.

In addition, the invention provides vectors comprising the polynucleotides/nucleic acid molecules of the invention.

Vectors are nucleic acid molecules that serve as vehicles for transferring (foreign) genetic material into cells. The term "vector" includes, but is not limited to, plasmids, viruses, cosmids, and artificial chromosomes. Typically, the engineered vector comprises an origin of replication, a multiple cloning site, and a selectable marker. The vector itself is typically a nucleotide sequence, usually a DNA sequence, which contains the insert (transgene) and a larger sequence that serves as the "backbone" of the vector. In addition to the transgenic insert and the backbone, modern vectors may also comprise other features: promoters, genetic markers, antibiotic resistance, reporter genes, targeting sequences, protein purification tags. Vectors, referred to as expression vectors (expression constructs), are used exclusively for expressing transgenes in target cells and usually have control sequences.

The term "control sequences" refers to DNA sequences necessary for the expression of an operably linked coding sequence in a particular host organism. Suitable control sequences for prokaryotes include, for example, promoters, optional operator sequences, and ribosome binding sites. Eukaryotic cells are known to utilize promoters, polyadenylation signals, and enhancers.

A nucleic acid is "operably linked" when it is placed into a functional relationship with another nucleic acid sequence. For example, DNA for a presequence or secretory leader is operably linked to DNA for a polypeptide if it is expressed as a preprotein that participates in the secretion of the polypeptide; a promoter or enhancer is operably linked to a coding sequence if it affects the transcription of the sequence; or operably linked to a coding sequence if the ribosome binding site is positioned to facilitate translation. In general, "operably linked" means that the DNA sequences being linked are contiguous and, in the case of secretory leader sequences, contiguous and in reading phase. However, enhancers need not be contiguous. Ligation is accomplished by ligation at appropriate restriction sites. If such sites are not present, synthetic oligonucleotide adaptors or linkers are used according to conventional practice.

"transfection" is the process of deliberately introducing nucleic acid molecules or polynucleotides (including vectors) into target cells. The term is used primarily for non-viral methods in eukaryotic cells. Transduction is generally used to describe the viral-mediated transfer of nucleic acid molecules or polynucleotides. Transfection of animal cells typically involves opening transient pores or "holes" in the cell membrane to allow uptake of the substance. Transfection may be performed using calcium phosphate, by electroporation, by cell extrusion, or by mixing cationic lipids with the substance that produces liposomes that fuse with the cell membrane and deposit their cargo inside.

The term "transformation" is used to describe the non-viral transfer of a nucleic acid molecule or polynucleotide (including vectors) into bacteria as well as non-animal eukaryotic cells (including plant cells). Thus, transformation is a genetic alteration of a bacterial or non-animal eukaryotic cell, which results from direct uptake from its environment through the cell membrane and subsequent incorporation of exogenous genetic material (nucleic acid molecules). The transformation may be effected manually. For transformation to occur, the cells or bacteria must be competent, which can occur as a time-limited response to environmental conditions (such as starvation and cell density).

In addition, the invention provides host cells transformed or transfected with the polynucleotide/nucleic acid molecules or vectors of the invention.

As used herein, the term "host cell" or "recipient cell" is intended to include any individual cell or cell culture that may be or has been a receptor for a vector, an exogenous nucleic acid molecule, and a polynucleotide encoding an antibody construct of the invention and/or a receptor for the antibody construct itself. The introduction of the corresponding substance into the cells is carried out by means of transformation, transfection or the like. The term "host cell" is also intended to include progeny or possible progeny of a single cell. Because certain modifications may occur in succeeding generations due to natural, accidental, or deliberate mutation, or due to environmental influences, such progeny may not, in fact, be identical to the parent cell (either in morphology or in genomic or total DNA complement), but are still included within the scope of the term as used herein. Suitable host cells include prokaryotic or eukaryotic cells, and also include, but are not limited to, bacteria, yeast cells, fungal cells, plant cells, and animal cells, such as insect cells and mammalian cells, e.g., murine, rat, cynomolgus, or human.

The antibody constructs of the invention may be produced in bacteria. After expression, the antibody constructs of the invention are separated from the E.coli cell paste in a soluble fraction and may be purified, for example, by affinity chromatography and/or size exclusion. The final purification can be carried out analogously to the purification of antibodies expressed, for example, in CHO cells.

In addition to prokaryotes, eukaryotic microorganisms such as filamentous fungi or yeast are suitable cloning or expression hosts for the antibody constructs of the invention. Saccharomyces cerevisiae or common baker's yeast is the most commonly used among lower eukaryotic host microorganisms. However, many other genera, species, and strains are generally available and useful herein, such as Schizosaccharomyces pombe (Schizosaccharomyces pombe); kluyveromyces hosts, such as kluyveromyces lactis (k.lactis), kluyveromyces fragilis (k.fragilis) (ATCC 12424), kluyveromyces bulgaricus (k.bulgaricus) (ATCC 16045), kluyveromyces willianus (k.wickramii) (ATCC 24178), kluyveromyces farinosus (k.waltii) (ATCC 56500), kluyveromyces drosophilus (k.hydro philum) (ATCC 36906), kluyveromyces thermotolerans (k.thermotolens), and kluyveromyces marxianus (k.marxianus); yarrowia (EP 402226); pichia pastoris (EP 183070); candida species; trichoderma reesei (Trichoderma reesei) (EP 244234); neurospora crassa (Neurospora crassa); schwanniomyces (Schwanniomyces), such as Schwanniomyces occidentalis (Schwanniomyces occidentalis); and filamentous fungi such as Neurospora, Penicillium, Tolypocladium, and Aspergillus hosts such as Aspergillus nidulans (A.nidulans) and Aspergillus niger (A.niger).

Host cells suitable for expressing the glycosylated antibody constructs of the present invention are derived from multicellular organisms. Examples of invertebrate cells include plant and insect cells. Numerous strains and variants of baculovirus from hosts such as Spodoptera frugiperda (caterpillar) (caterpillars), Aedes aegypti (mosquitoes), Aedes albopictus (mosquitoes), Drosophila melanogaster (Drosophila melanogaster) and Bombyx mori (Bombyx mori) have been identified, as well as corresponding permissive insect host cells. Various virus strains for transfection are publicly available, such as the L-1 variant of Autographa californica (NPV) and the Bm-5 strain of Bombyx mori (Bombyx mori) NPV, and such viruses can be used as the virus herein according to the present invention, particularly for transfecting Spodoptera frugiperda cells.

Plant cell cultures of cotton, corn, potato, soybean, petunia, tomato, arabidopsis and tobacco may also be used as hosts. Cloning and expression vectors useful for the production of proteins in plant cell culture are known to those skilled in the art. See, e.g., Hiatt et al, Nature (1989) 342: 76-78; owen et al (1992) Bio/Technology 10: 790 and 794; artsaecko et al (1995) The Plant J8: 745-750; and Fecker et al (1996) Plant Mol Biol 32: 979-986.

However, vertebrate cells are of most interest, and propagation of vertebrate cells in culture (tissue culture) has become a routine procedure. Examples of useful mammalian host cell lines are monkey kidney CV1 cell line transformed with SV40 (COS-7, ATCC CRL 1651); human embryonic kidney cell lines (293 or 293 cells subcloned for growth in suspension culture, Graham et al, J.Gen Virol.36: 59 (1977)); baby hamster kidney cells (BHK, ATCC CCL 10); chinese hamster ovary cells/-DHFR (CHO, Urlaub et al, Proc. Natl. Acad. Sci. USA 77: 4216 (1980)); mouse testicular support cells (TM4, Mather, biol. reprod.23: 243-251 (1980)); monkey kidney cells (CVI ATCC CCL 70); vero cells (VERO-76, ATCC CRL 1587); human cervical cancer cells (HELA, ATCC CCL 2); canine kidney cells (MDCK, ATCC CCL 34); buffalo rat (buffalo rat) hepatocytes (BRL 3A, ATCC CRL 1442); human lung cells (W138, ATCC CCL 75); human hepatocytes (Hep G2, 14138065); mouse mammary tumor (MMT 060562, ATCC CCL 51); TRI cells (Mather et al, Annals N.Y. Acad.Sci. (1982) 383: 44-68); MRC 5 cells; FS4 cells; and a human hepatoma line (Hep G2).

In another embodiment, the invention provides a method for producing an antibody construct of the invention, the method comprising culturing a host cell of the invention under conditions that allow expression of the antibody construct of the invention, and recovering the produced antibody construct from the culture.

As used herein, the term "culturing" refers to the in vitro maintenance, differentiation, growth, proliferation and/or propagation of cells in culture medium under appropriate conditions. The term "expression" includes any step involved in the production of the antibody constructs of the invention, including but not limited to transcription, post-transcriptional modification, translation, post-translational modification, and secretion.

When recombinant techniques are used, the antibody construct may be produced intracellularly, in the periplasmic space, or directly secreted into the culture medium. If the antibody construct is produced intracellularly, as a first step, particulate debris (host cells or lysed fragments) are removed, for example by centrifugation or ultrafiltration. Carter et al, Bio/Technology 10: 163-167(1992) describes a method for isolating antibodies secreted into the periplasmic space of E.coli. Briefly, the cell paste was thawed in the presence of sodium acetate (pH 3.5), EDTA, and phenylmethylsulfonyl fluoride (PMSF) within about 30 minutes. Cell debris can be removed by centrifugation. When the antibody is secreted into the culture medium, the supernatant from such an expression system is typically first concentrated using a commercially available protein concentration filter (e.g., Amicon or Millipore Pellicon ultrafiltration unit). Protease inhibitors such as PMSF may be included in any of the above steps to inhibit proteolysis, and antibiotics may be included to prevent the growth of adventitious contaminants.

The antibody construct of the invention prepared from the host cell may be recovered or purified using, for example, hydroxyapatite chromatography, gel electrophoresis, dialysis, and affinity chromatography. Depending on the antibody to be recovered, other techniques for protein purification may also be used, such as fractionation on ion exchange columns, ethanol precipitation, reverse phase HPLC, chromatography on silica, heparin SEPHAROSE^TMChromatography on an anion or cation exchange resin (such as a polyaspartic acid column), chromatofocusing, SDS-PAGE and ammonium sulfate precipitation. When the antibody construct of the invention comprises a CH3 domain, Bakerbond ABX resin (j.t. baker, phillips burg, NJ) is useful for purification.

Affinity chromatography is a preferred purification technique. The matrix to which the affinity ligand is attached is most commonly agarose, but other matrices are also useful. Mechanically stable matrices such as controlled pore glass or poly (styrene-divinyl) benzene allow faster flow rates and shorter processing times than can be achieved with agarose.

Furthermore, the invention provides a pharmaceutical composition comprising the antibody construct of the invention or produced according to the method of the invention. For the pharmaceutical composition of the present invention, it is preferred that the homogeneity of the antibody construct is ≥ 80%, more preferably ≥ 81%, ≥ 82%, ≥ 83%, ≥ 84% or ≥ 85%, even more preferably ≥ 86%, > 87%, > 88%, > 89% or ≥ 90%, even more preferably ≥ 91%, > 92%, > 93%, > 94% or ≥ 95%, most preferably ≥ 96%, > 97%, > 98% or ≥ 99%.

As used herein, the term "pharmaceutical composition" relates to a composition suitable for administration to a patient, preferably a human patient. Particularly preferred pharmaceutical compositions of the invention preferably comprise one or more antibody constructs of the invention in a therapeutically effective amount. Preferably, the pharmaceutical composition further comprises suitable formulations of one or more (pharmaceutically effective) carriers, stabilizers, excipients, diluents, solubilizers, surfactants, emulsifiers, preservatives and/or adjuvants. Acceptable ingredients of the composition are preferably non-toxic to recipients at the dosages and concentrations employed. The pharmaceutical compositions of the present invention include, but are not limited to, liquid, frozen and lyophilized compositions.

The compositions of the present invention may comprise a pharmaceutically acceptable carrier. Generally, as used herein, "pharmaceutically acceptable carrier" means any and all aqueous and non-aqueous solutions, sterile solutions, solvents, buffers such as Phosphate Buffered Saline (PBS) solutions, water, suspensions, emulsions such as oil/water emulsions, various types of wetting agents, liposomes, dispersion media and coatings that are compatible with pharmaceutical administration, particularly with parenteral administration. The use of such media and agents in pharmaceutical compositions is well known in the art, and compositions containing such carriers can be formulated by well known conventional methods.

Certain embodiments provide pharmaceutical compositions comprising an antibody construct of the invention and additionally one or more excipients, such as the excipients illustratively described in this section and elsewhere herein. In this regard, excipients may be used in the present invention for various purposes, such as to modulate the physical, chemical or biological properties of the formulation, such as to modulate viscosity and/or to be used in the methods of the present invention to improve effectiveness and/or to stabilize such formulations and processes against degradation and spoilage due to stresses occurring, for example, during and after manufacture, transport, storage, preparation prior to use, application.

In certain embodiments, the PHARMACEUTICAL compositions may contain formulation materials for the purpose of altering, maintaining or maintaining, for example, the pH, osmotic pressure, viscosity, clarity, color, isotonicity, odor, sterility, stability, dissolution or release rate, adsorption or permeation of the composition (see REMINGTON' S PHARMACEUTICAL scientific SCIENCES, 18 th edition, (a.r. genmo, ed.), 1990, Mack Publishing Company). In such embodiments, suitable formulation materials may include, but are not limited to:

● amino acids such as glycine, alanine, glutamine, asparagine, threonine, proline, 2-phenylalanine, including charged amino acids, preferably lysine, lysine acetate, arginine, glutamic acid and/or histidine

● antimicrobial agents such as antibacterial and antifungal agents

● antioxidants such as ascorbic acid, methionine, sodium sulfite or sodium bisulfite;

● buffers, buffer systems and buffers for maintaining the composition at physiological pH or at a slightly lower pH, typically in the range of about 5 to about 8 or 9; examples of buffering agents are borate, bicarbonate, Tris-HCl, citrate, phosphate or other organic acids, succinate, phosphate, histidine and acetate; for example, Tris buffer at about pH 7.0-8.5 or acetate buffer at about pH 4.0-5.5;

● non-aqueous solvents such as propylene glycol, polyethylene glycol, vegetable oils such as olive oil, and injectable organic esters such as ethyl oleate;

● aqueous carrier including water, alcohol/water solution, emulsion or suspension, including saline and buffer medium;

● biodegradable polymers such as polyesters;

● bulking agents such as mannitol or glycine;

● chelating agents such as ethylenediaminetetraacetic acid (EDTA);

● isotonic and absorption delaying agent;

● complexing agents such as caffeine, polyvinylpyrrolidone, beta-cyclodextrin or hydroxypropyl-beta-cyclodextrin

● a filler;

● a monosaccharide; a disaccharide; and other carbohydrates (such as glucose, mannose, or dextrins); the carbohydrate may be a non-reducing sugar, preferably trehalose, sucrose, octasulfate, sorbitol or xylitol;

● (low molecular weight) protein, polypeptide or proteinaceous carrier, such as human or bovine serum albumin, gelatin or immunoglobulin, preferably of human origin;

● coloring and flavoring agents;

● sulfureous reducing agents, such as glutathione, lipoic acid, sodium thioglycolate, thioglycerol, [ alpha ] -monothioglycerol and sodium thiosulfate

● a diluent;

● an emulsifier;

● hydrophilic polymers such as polyvinylpyrrolidone

● salt-forming counterions such as sodium;

● preservatives such as antimicrobials, antioxidants, chelating agents, inert gases and the like; examples are: benzalkonium chloride, benzoic acid, salicylic acid, thimerosal, phenyl ethanol, methyl paraben, propyl paraben, chlorhexidine, sorbic acid, or hydrogen peroxide).

● metal complexes such as zinc-protein complexes;

● solvents and cosolvents (such as glycerol, propylene glycol or polyethylene glycol);

● sugars and sugar alcohols, such as trehalose, sucrose, octasulfate, mannitol, sorbitol or xylitol stachyose, mannose, sorbose, xylose, ribose, inositol, galactose, lactitol, ribitol, inositol, galactitol, glycerol, cyclic alcohols (e.g., inositol), polyethylene glycol; and polyhydric sugar alcohols;

● suspending agent;

● surfactants or wetting agents such as pluronics, PEG, sorbitan esters, polysorbates such as polysorbate 20, polysorbate, triton, tromethamine, lecithin, cholesterol, tyloxapol; the surfactant may be a detergent, preferably having a molecular weight > 1.2KD and/or a polyether, preferably having a molecular weight > 3 KD; non-limiting examples of preferred detergents are tween 20, tween 40, tween 60, tween 80 and tween 85; non-limiting examples of preferred polyethers are PEG 3000, PEG 3350, PEG 4000 and PEG 5000;

● stability enhancers such as sucrose or sorbitol;

● tonicity enhancing agents such as alkali metal halides, preferably sodium or potassium chloride, mannitol sorbitol;

● parenteral delivery vehicles including sodium chloride solution, ringer's dextrose, dextrose and sodium chloride, lactated ringer's solution or fixed oils;

● intravenous delivery vehicles including fluid and nutritional supplements, electrolyte supplements (such as ringer's glucose based supplements).

It will be apparent to those skilled in the art that different ingredients of a pharmaceutical composition (e.g., those listed above) may have different effects, e.g., an amino acid may act as a buffer, stabilizer, and/or antioxidant; mannitol may act as a bulking agent and/or tonicity enhancer; sodium chloride may act as a delivery vehicle and/or tonicity enhancer; etc. of

It is envisaged that the compositions of the invention may comprise, in addition to the polypeptides of the invention defined herein, other biologically active agents, depending on the intended use of the composition. Such agents may be drugs acting on the gastrointestinal system, drugs acting as cytostatics, drugs preventing hyperuricemia, drugs inhibiting immune responses (e.g., corticosteroids), drugs modulating inflammatory responses, drugs acting on the circulatory system and/or agents known in the art such as cytokines. It is also envisaged that the antibody constructs of the invention are used in co-therapy, i.e. in combination with another anti-cancer drug.

In certain embodiments, the optimal pharmaceutical composition will be determined by one of skill in the art based on, for example, the intended route of administration, delivery form, and desired dosage. See, e.g., REMINGTON' S PHARMACEUTICAL SCIENCES, supra. In certain embodiments, such compositions may affect the physical state, stability, rate of in vivo release, and rate of in vivo clearance of the antibody constructs of the invention. In certain embodiments, the primary vehicle or carrier in the pharmaceutical composition may be aqueous or non-aqueous in nature. For example, a suitable vehicle or carrier may be water for injection, physiological saline solution, or artificial cerebrospinal fluid, possibly supplemented with other substances common in compositions for parenteral administration. Neutral buffered saline or saline mixed with serum albumin are further exemplary vehicles. In certain embodiments, the antibody constructs of the compositions of the present invention may be prepared for storage by mixing the selected composition of the desired purity with an optional formulation (REMINGTON' S PHARMACEUTICAL SCIENCES, supra) in the form of a lyophilized cake or an aqueous solution. In addition, in certain embodiments, the antibody constructs of the invention may be formulated as a lyophilizate using a suitable excipient such as sucrose.

When parenteral administration is contemplated, the therapeutic compositions for use in the present invention may be provided in the form of a pyrogen-free, parenterally acceptable aqueous solution comprising the desired antibody construct of the invention in a pharmaceutically acceptable vehicle. A particularly suitable vehicle for parenteral injection is sterile distilled water, wherein the antibody constructs of the invention are formulated as sterile isotonic solutions for proper storage. In certain embodiments, preparation may involve formulating an agent that can provide controlled or sustained release of a product that can be delivered by reservoir injection, such as injectable microspheres, bioerodible particles, polymeric compounds (such as polylactic acid or polyglycolic acid), beads, or liposomes, with the desired molecule. In certain embodiments, hyaluronic acid may also be used, which has the effect of promoting circulation duration. In certain embodiments, the implantable drug delivery device can be used to introduce a desired antibody construct.

Additional pharmaceutical compositions will be apparent to those skilled in the art, including formulations involving the antibody constructs of the invention in sustained or controlled delivery/release formulations. Techniques for formulating various other sustained or controlled delivery means, such as liposome carriers, bioerodible microparticles or porous beads, and depot injections, are also known to those skilled in the art. See, for example, International patent application No. PCT/US93/00829, which describes controlled release of porous polymeric microparticles for delivery of pharmaceutical compositions. Sustained release formulations may include a semipermeable polymer matrix in the form of a shaped article (e.g., a film or microcapsule). Sustained release matrices may include polyesters, hydrogels, polylactides (as disclosed in U.S. Pat. No. 3,773,919 and European patent application publication No. EP 058481), copolymers of L-glutamic acid and gamma-ethyl-L-glutamic acid (Sidman et al, 1983, Biopolymers 2: 547-556), poly (2-hydroxyethyl-methacrylate) (Langer et al, 1981, J.biomed.Mater.Res.15: 167-277 and Langer, 1982, chem.Tech.12: 98-105), ethylene vinyl acetate (Langer et al, 1981, supra) or poly-D (-) -3-hydroxybutyric acid (European patent application publication No. EP 133,988). Sustained release compositions may also include liposomes that can be prepared by any of several methods known in the art. See, e.g., Eppstein et al, 1985, proc.natl.acad.sci.u.s.a.82: 3688-; european patent application publications EP 036,676, EP 088,046 and EP 143,949.

The antibody construct may also be encapsulated in a microcapsule package, a colloidal drug delivery system (e.g., liposomes, albumin microspheres, microemulsions, nanoparticles, and nanocapsules), or a macroemulsion, for example, prepared by coacervation techniques or by interfacial polymerization (e.g., hydroxymethylcellulose or gelatin-microcapsules and poly (methylmethacylate) microcapsules, respectively). Such techniques are disclosed in Remington's Pharmaceutical Sciences, 16 th edition, Oslo, a. editor, (1980).

Pharmaceutical compositions for in vivo administration are typically provided as sterile formulations. Sterilization may be accomplished by filtration through sterile filtration membranes. When the composition is lyophilized, sterilization using this method can be performed before or after lyophilization and reconstitution. Compositions for parenteral administration may be stored in lyophilized form or in solution. Parenteral compositions are typically placed into a container having a sterile access port (e.g., an intravenous solution bag or vial having a stopper pierceable by a hypodermic injection needle).

Another aspect of the invention comprises a self-buffering antibody construct of the formulation of the invention useful as a pharmaceutical composition, as described in International patent application WO 06138181A2(PCT/US 2006/022599). Various descriptions are available regarding protein stabilization and formulation materials and methods for use in this regard, such as Arakawa et al, "Solvent interactions in pharmaceutical formulations," Pharm Res.8 (3): 285-91 (1991); kendrick et al, RATIONAL DESIGN OF STABLE PROTEIN FORMATIONS: THEORY AND PRACTICE, Carpenter and Manning, editors Pharmaceutical biotechnology.13: 61-84(2002), "physiological stabilization of proteins in aqueous solution" and Randolph et al, "Surfactant-protein interactions", Pharm Biotechnol.13: 159-75(2002), see in particular the section on excipients and methods relating to self-buffering protein formulations according to the invention, especially for protein pharmaceutical products and methods for veterinary and/or human medical use.

According to certain embodiments of the invention, salts may be used, for example, to adjust the ionic strength and/or isotonicity of the formulation and/or to improve the solubility and/or physical stability of the protein or other ingredients of the composition according to the invention. As is well known, ions can stabilize the native state of proteins by binding to charged residues on the surface of the protein, by shielding the charged polar groups in the protein and reducing the strength, attraction, and repulsion of their electrostatic interactions. The ions may also stabilize the denatured state of the protein by binding specifically to the protein's altered peptide linkage (- -CONH). In addition, ionic interactions with charged polar groups in proteins can also reduce intermolecular electrostatic interactions, thereby preventing or reducing aggregation and insolubility of proteins.

Ionic species differ significantly in their effect on proteins. A number of categorical orderings of ions and their effect on proteins that can be used to formulate pharmaceutical compositions according to the present invention have been developed. One example is the hofmeister series, which fractionates ionic and polar non-ionic solutes by their effect on the conformational stability of proteins in solution. The stabilizing solute is called "lyophilic". Destabilizing solutes are referred to as "chaotropes". Kosmotropic agents are typically used at high concentrations (e.g., > 1 molar ammonium sulfate) to precipitate proteins from solution ("salting out"). Chaotropic agents are commonly used to denature and/or solubilize proteins ("salting-in"). The relative effectiveness of the ion pairs "salting in" and "salting out" defines their position in the hofmeister series.

Free amino acids may be used as bulking agents, stabilizers and antioxidants in the antibody construct formulations of the invention according to various embodiments of the invention, as well as other standard uses. Lysine, proline, serine and alanine may be used to stabilize the protein in the formulation. Glycine was used for lyophilization to ensure correct cake structure and properties. Arginine can be used to inhibit protein aggregation in both liquid and lyophilized formulations. Methionine can be used as an antioxidant.

Polyols include sugars such as mannitol, sucrose and sorbitol, and polyols such as, for example, glycerol and propylene glycol, as well as polyethylene glycol (PEG) and related materials for the purposes discussed herein. The polyol is lyophilic. They are useful stabilizers in liquid and lyophilized formulations to protect proteins from physical and chemical degradation processes. Polyols are also used to adjust the tonicity of the formulation. A polyol useful in selected embodiments of the present invention is mannitol, which is typically used to ensure structural stability of the filter cake in lyophilized formulations. Which ensures structural stability of the filter cake. It is usually used with a lyoprotectant such as sucrose. Sorbitol and sucrose are preferred agents among those used to adjust tonicity and as stabilizers to prevent freeze-thaw stress during transport or when preparing briquettes in the manufacturing process. Reducing sugars (which contain free aldehyde or ketone groups), such as glucose and lactose, can saccharify surface lysine and arginine residues. Therefore, they are generally not among the preferred polyols for use according to the present invention. In addition, sugars that form such reactive species, such as sucrose (which hydrolyses to fructose and glucose under acidic conditions and thus leads to saccharification), are also not among the preferred polyols of the present invention in this regard. PEG is useful for stabilizing proteins and as a cryoprotectant, and may be used in this regard in the present invention.

Embodiments of the antibody construct of the formulations of the invention further comprise a surfactant. Protein molecules can be readily adsorbed on surfaces and denatured and subsequently aggregated at gas-liquid, solid-liquid and liquid-liquid interfaces. These effects are generally inversely proportional to protein concentration. These deleterious interactions are generally inversely proportional to protein concentration and are often exacerbated by physical agitation, such as that produced during the transport and handling of the product. Surfactants are commonly used to prevent, minimize or reduce surface adsorption. In this regard, surfactants useful in the present invention include polysorbate 20, polysorbate 80, other fatty acid esters of sorbitan polyethoxylate, and poloxamer 188. Surfactants are also commonly used to control protein conformational stability. The use of surfactants in this regard is protein specific, as any given surfactant will generally stabilize some proteins and destabilize others.

Polysorbates are susceptible to oxidative degradation and typically contain a sufficient amount of peroxide at the time of supply to cause oxidation of the side chains of protein residues, particularly methionine. Therefore, polysorbate should be used with caution and should be used at its lowest effective concentration. In this regard, polysorbates illustrate the general rule that excipients should be used at their lowest effective concentration.

Embodiments of the antibody constructs of the invention further comprise one or more antioxidants. To some extent, detrimental oxidation of proteins in pharmaceutical formulations can be prevented by maintaining appropriate levels of ambient oxygen and temperature, and by avoiding exposure to light. Antioxidant excipients may also be used to prevent oxidative degradation of proteins. Among the useful antioxidants in this regard are reducing agents, oxygen/free radical scavengers, and chelating agents. The antioxidant used in the therapeutic protein formulation according to the invention is preferably water soluble and retains its activity throughout the shelf life of the product. In this respect, EDTA is a preferred antioxidant according to the present invention. Antioxidants can destroy proteins. For example, reducing agents, such as glutathione in particular, can disrupt intramolecular disulfide linkages. Thus, the antioxidant used in the present invention is selected to, inter alia, eliminate or substantially reduce the possibility of itself destroying the protein in the formulation.

Formulations according to the invention may include metal ions that are protein cofactors and are necessary to form protein coordination complexes, such as zinc, which is necessary to form certain insulin suspensions. Metal ions can also inhibit some processes that degrade proteins. However, metal ions also catalyze physical and chemical processes that degrade proteins. Magnesium ions (10-120mM) can be used to inhibit the isomerization of aspartic acid to isoaspartic acid. Ca⁺²Ions (up to 100mM) improve the stability of human DNase. However, Mg⁺²、Mn⁺²And Zn⁺²Destabilizes the rhDN enzyme. Similarly, Ca⁺²And Sr⁺²Can stabilize factor VIII, which can be Mg⁺²、Mn⁺²And Zn⁺²、Cu⁺²And Fe⁺²Destabilization and passing Al⁺³The ions increase their aggregation.

Embodiments of the antibody construct of the formulations of the invention further comprise one or more preservatives. Preservatives are necessary when developing multi-dose parenteral formulations, which development involves more than one extraction from the same container. Their primary function is to inhibit microbial growth and ensure sterility of the product throughout the shelf-life or useful life of the pharmaceutical product. Common preservatives include benzyl alcohol, phenol and m-cresol. While preservatives have a long history of use in small molecule parenteral applications, the development of protein formulations containing preservatives can be challenging. Preservatives almost always have an unstable effect on the protein (aggregation) and this has become a major factor limiting its use in multi-dose protein formulations. To date, most protein drugs have been formulated for single use only. However, when multiple dose formulations are possible, they have the additional advantage of providing patient convenience and increasing marketability. A good example is the multi-dose formulation of human growth hormone (hGH), wherein the development of a preservative formulation has led to the commercialization of a more convenient, multi-purpose injection pen presentation. There are at least four such pen devices currently on the market that contain a preserved formulation of hGH. Nodroxin (Norditropin) (liquid, Novo Nordisk), Nutropin AQ (liquid, Genentech) and Genotropin (lyophilized-two-chamber cartridge, Pharmacia & Upjohn) contain phenol, while somatrix (eli lilly) is formulated with m-cresol. Several aspects need to be considered in the formulation and development of preserved dosage forms. The effective preservative concentration in the drug product must be optimized. This requires testing the concentration range of a given preservative in the dosage form that confers antimicrobial efficacy without compromising protein stability.

As can be expected, the development of a liquid formulation containing a preservative is more challenging than a lyophilized formulation. The freeze-dried product may be lyophilized without a preservative and reconstituted at the time of use with a diluent containing a preservative. This shortens the time of contact of the preservative with the protein, significantly minimizing the associated stability risks. For liquid formulations, preservative efficacy and stability should be maintained throughout the shelf life of the product (about 18 to 24 months). Importantly, the effectiveness of the preservative should be demonstrated in the final formulation containing the active drug and all the excipient components.

The antibody constructs disclosed herein can also be formulated as immunoliposomes. "liposomes" are vesicles composed of various types of lipids, phospholipids, and/or surfactants that can be used to deliver drugs to mammals. The components of liposomes are typically arranged in a bilayer, similar to the lipid arrangement of biological membranes. Liposomes containing the antibody construct are prepared by methods known in the art, such as Epstein et al, proc.natl.acad.sci.usa, 82: 3688 (1985); hwang et al, proc.natl acad.sci.usa, 77: 4030 (1980); U.S. patent nos. 4,485,045 and 4,544,545; and W097/38731. Liposomes with extended circulation time are disclosed in U.S. patent No. 5,013,556. Particularly useful liposomes can be produced by reverse phase evaporation using a lipid composition comprising phosphatidylcholine, cholesterol and PEG-derivatized phosphatidylethanolamine (PEG-PE). The liposomes are extruded through a filter having a defined pore size to produce liposomes having a desired diameter. Such as Martin et al, j.biol.chem.257: 286-288(1982), the Fab' fragment of the antibody construct of the invention was conjugated to liposomes by a disulfide exchange reaction. The chemotherapeutic agent is optionally contained within the liposome. See Gabizon et al J.national Cancer Inst.81(19)1484 (1989).

Once the pharmaceutical composition is formulated, it may be stored in sterile vials as a solution, suspension, gel, emulsion, solid, crystal, or as a dehydrated or lyophilized powder. Such formulations may be stored in a ready-to-use form or in a form that is reconstituted prior to administration (e.g., lyophilized).

The biological activity of the pharmaceutical compositions as defined herein can be determined, for example, by a cytotoxicity assay, as described in the examples in WO 99/54440 or by Schlereth et al (Cancer immunol. immunother.20(2005), 1-12). As used herein, "efficacy" or "in vivo efficacy" refers to the response to treatment by a pharmaceutical composition of the invention using, for example, standardized NCI response criteria. The success or in vivo efficacy of a treatment using the pharmaceutical composition of the invention refers to the effectiveness of the composition for its intended purpose, i.e., the ability of the composition to elicit its desired effect, i.e., the depletion of pathological cells (e.g., tumor cells). In vivo efficacy can be monitored by established standard methods for the respective disease entity, including but not limited to, white blood cell count, differential, fluorescence activated cell sorting, bone marrow aspiration. In addition, various disease-specific clinical chemistry parameters and other established standard methods can be used. In addition, computer-assisted tomography, X-ray, nuclear magnetic resonance tomography (e.g., for response assessment based on national institute of cancer standards [ Cheson BD, Horning SJ, Coiffier B, Shipp MA, Fisher RI, comparators JM, Lister TA, Vose J, Grillo-Lopez A, Hagenbeek A, Cabanillales F, Klippensten D, Hiddemann W, Castellino R, Harris NL, Arimitage JO, Carter W, Hoppe R, Canelos GP. report of interfacial work to stable diameter sensitivity criterion for non-Hodgkin's lymphoma. NCI Spondred International Working group, J Clin 4; 12417 month for biopsy, positron emission/emission tomography, tissue emission/emission tomography, positron emission tomography (PET/emission), and various lymphoma-specific clinical chemistry parameters (e.g., lactate dehydrogenase) and other established standard methods can be used.

Another major challenge in the development of drugs such as the pharmaceutical compositions of the present invention is the modulation of predictable pharmacokinetic properties. To this end, pharmacokinetic profiles of drug candidates, i.e. profiles of pharmacokinetic parameters that affect the ability of a particular drug to treat a given condition, may be established. Pharmacokinetic parameters of drugs that affect the ability of the drug to treat certain disease entities include, but are not limited to: half-life, volume of distribution, hepatic first-pass metabolism and serum binding. The efficacy of a given drug may be affected by each of the parameters described above.

By "half-life" is meant the time during which 50% of the administered drug is eliminated by biological processes such as metabolism, excretion, and the like. "hepatic first pass metabolism" refers to the tendency of a drug to be metabolized when it first contacts the liver (i.e., during its first pass through the liver). By "distribution volume" is meant the degree of retention of the drug in different compartments within the body, such as, for example, intracellular and extracellular spaces, tissues and organs, etc., as well as the distribution of the drug within these compartments. By "serum binding" is meant the tendency of a drug to interact with and bind to a serum protein, such as albumin, resulting in a reduction or loss of the biological activity of the drug.

Pharmacokinetic parameters also include the bioavailability, lag time (Tlag), Tmax, absorption, more episodes (more onset) and/or Cmax of a given amount of administered drug. By "bioavailability" is meant the amount of drug in the blood compartment. By "lag time" is meant the time delay between administration of the drug and its detection and measurement in blood or plasma. "Tmax" is the time required for the drug to reach maximum plasma concentration, and "Cmax" is the maximum plasma concentration obtained for a given drug. The time to reach the blood or tissue concentration of the drug required for its biological action is influenced by all parameters. Pharmacokinetic parameters for bispecific antibody constructs that exhibit cross-species specificity are also set forth in, for example, Schlereth et al publications (Cancer immunol. immunother.20(2005), 1-12) (which can be determined in non-chimpanzee primates in preclinical animal testing as outlined above).

One embodiment provides an antibody construct of the invention or an antibody construct produced according to the method of the invention for use in the prevention, treatment or amelioration of a tumor or cancer disease or a metastatic cancer disease.

According to a preferred embodiment of the invention, the tumor or cancer disease is a solid tumor disease.

The formulations described herein may be used as pharmaceutical compositions for treating, ameliorating and/or preventing a pathological medical condition as described herein in a patient in need thereof. The term "treatment" refers to both therapeutic treatment and prophylactic or preventative measures. Treatment includes applying or administering the formulation to the body, isolated tissue or cells of a patient suffering from a disease/disorder, a symptom of a disease/disorder, or having a predisposition for a disease/disorder, with the purpose of curing, healing, alleviating, altering, remediating, ameliorating, augmenting, or affecting the disease, the symptoms of the disease, or the predisposition for the disease.

The term "ameliorating" as used herein refers to any improvement in the disease state of a patient suffering from a tumor or cancer or metastatic cancer as described below by administering to a subject in need thereof an antibody construct according to the invention. Such an improvement may also be seen as a slowing or stopping of the progression of the tumor or cancer or metastatic cancer in the patient. As used herein, the term "prevention" means that a patient is prevented from developing or relapsing from a tumor or cancer or metastatic cancer as described below by administering an antibody construct according to the invention to a subject in need thereof.

The term "disease" refers to any condition that would benefit from treatment with an antibody construct or pharmaceutical composition described herein. This includes chronic and acute conditions or diseases, including those pathological conditions that predispose a mammal to the disease.

A "tumor" is an abnormal growth of tissue, usually but not always forming a lump. When a lump is also formed, it is often referred to as a "tumor". A neoplasm or tumor may be benign, possibly malignant (pre-cancerous) or malignant. Malignant neoplasms are commonly referred to as cancers. They often invade and destroy surrounding tissue and may form metastases, i.e. they spread to other parts, tissues or organs of the body. Thus, the term "metastatic cancer" includes metastasis to other tissues or organs compared to the tissues or organs other than the original tumor. Lymphomas and leukemias are lymphomas. For the purposes of the present invention, they are also encompassed by the terms "tumor" or "cancer".

In a preferred embodiment of the invention, the tumor or cancer disease is a solid tumor disease, and the metastatic cancer disease may be derived from any of the above-mentioned diseases.

In connection with the present invention, preferably the tumor or cancer disease is selected from the group consisting of: glioblastoma, astrocytoma, medulloblastoma, breast cancer, non-small cell lung cancer, ovarian cancer, prostate cancer, central nervous system cancer. More preferably, the tumor or cancer disease is glioblastoma multiforme (GBM) or anaplastic astrocytoma. Metastatic cancer diseases may be derived from any of the above-mentioned diseases.

The invention also provides a method for treating or ameliorating a tumor or cancer disease or a metastatic cancer disease comprising the step of administering to a subject in need thereof an antibody construct of the invention or produced according to a method of the invention.

The terms "subject in need thereof" or "those in need of treatment" include those already having the disorder as well as those in which the disease is to be prevented. Subjects or "patients" in need thereof include human and other mammalian subjects receiving prophylactic or therapeutic treatment.

The antibody constructs of the invention will generally be designed for a particular route and method of administration, a particular dosage and frequency of administration, a particular treatment for a particular disease, a range of bioavailability and persistence, and the like. The materials of the composition are preferably formulated at a concentration that is acceptable for the site of application.

Formulations and compositions may therefore be designed in accordance with the present invention for delivery by any suitable route of administration. In the context of the present invention, routes of administration include, but are not limited to

● topical routes (such as epidermal, inhalation, nasal, ocular, otic/otic, vaginal, transmucosal);

● intestinal route (such as oral, gastrointestinal, sublingual, sublabial, buccal, rectal); and

● parenteral route (such as intravenous, intra-arterial, intraosseous, intramuscular, intracerebral, intracerebroventricular, epidural, intrathecal, subcutaneous, intraperitoneal, extraamniotic, intraarticular, intracardiac, intradermal, intralesional, intrauterine, intravesical, intravitreal, transdermal, intranasal, transmucosal, intrasynovial, intraluminal).

The pharmaceutical compositions and antibody constructs of the invention are particularly suitable for parenteral administration, e.g., subcutaneous or intravenous delivery, e.g., by injection such as bolus injection, or by infusion such as continuous infusion. The pharmaceutical composition may be administered using a medical device. Examples of medical devices for administering pharmaceutical compositions are described in U.S. Pat. nos. 4,475,196, 4,439,196, 4,447,224, 4,447,233, 4,486,194, 4,487,603, 4,596,556, 4,790,824, 4,941,880, 5,064,413, 5,312,335, 5,312,335, 5,383,851 and 5,399,163.

In particular, the present invention provides for uninterrupted administration of suitable compositions. By way of non-limiting example, uninterrupted or substantially uninterrupted, i.e., continuous administration, may be achieved by a small pump system worn by the patient for metering the inflow of the therapeutic agent into the patient. Pharmaceutical compositions comprising the antibody constructs of the invention can be administered by using the pump system. Such pump systems are generally known in the art and typically rely on periodic replacement of a cartridge containing the therapeutic agent to be infused. When changing cartridges in such pump systems, the otherwise uninterrupted flow of therapeutic agent into the patient may be temporarily interrupted. In this case, the administration phase before cartridge replacement and the administration phase after cartridge replacement are still to be considered within the meaning of the pharmaceutical means and method of the present invention and together constitute one "uninterrupted administration" of such therapeutic agents.

Continuous or uninterrupted administration of the antibody construct of the invention may be intravenous or subcutaneous by means of a fluid delivery device or a mini-pump system comprising a fluid drive mechanism for driving fluid out of the reservoir and an actuation mechanism for actuating the drive mechanism. A pump system for subcutaneous administration may include a needle or cannula for penetrating the skin of a patient and delivering a suitable composition into the patient. The pump system may be secured or attached directly to the patient's skin independent of the vein, artery or blood vessel, allowing direct contact between the pump system and the patient's skin. The pump system may be attached to the skin of the patient for 24 hours to several days. The pump system may be of a small size and have a reservoir for a small volume. As a non-limiting example, the volume of the reservoir for a suitable pharmaceutical composition to be administered may be between 0.1 and 50 ml.

Continuous application may also be applied transdermally by means of a patch that is worn on the skin and replaced at intervals. Those skilled in the art are aware of patch systems suitable for drug delivery for this purpose. It is noted that transdermal administration is particularly suitable for uninterrupted administration, since the replacement of a new second patch after the first patch has been used up can advantageously be done simultaneously, for example by placing the second patch next to the skin surface of the used up first patch and subsequently removing the used up first patch. No flow interruption or battery failure problems occur.

If the pharmaceutical composition has been lyophilized, the lyophilized material is first reconstituted in a suitable liquid prior to administration. The lyophilized material may be reconstituted in, for example, bacteriostatic water for injection (BWFI), physiological saline, Phosphate Buffered Saline (PBS), or the same formulation as the protein prior to lyophilization.

The compositions of the invention can be administered to a subject at an appropriate dose, which can be determined, for example, by a dose escalation study by administering to a non-chimpanzee primate (e.g., cynomolgus monkey) an escalating dose of an antibody construct of the invention exhibiting cross-species specificity as described herein. As mentioned above, the antibody constructs of the invention exhibiting cross-species specificity as described herein can advantageously be used in the same form for preclinical testing in non-chimpanzee primates, as well as for use as human medicaments. The dosage regimen will be determined by the attending physician and clinical factors. As is well known in the medical arts, the dosage for any one patient depends on many factors, including the patient's size, body surface area, age, the particular compound to be administered, sex, time and route of administration, general health, and other drugs being administered concurrently.

The term "effective dose" or "effective pharmaceutical amount" is defined as an amount sufficient to achieve, or at least partially achieve, a desired effect. The term "therapeutically effective dose" is defined as an amount sufficient to cure or at least partially arrest a disease and its complications in a patient already suffering from the disease. The amount or dose effective for this use will depend on the condition to be treated (the indication), the antibody construct delivered, the therapeutic background and objectives, the severity of the disease, previous therapy, the patient's clinical history and response to the therapeutic agent, the route of administration, the patient's size (body weight, body surface or organ size) and/or condition (age and general health), and the general state of the patient's own immune system. The appropriate dosage may be adjusted at the discretion of the attending physician so that the dosage may be administered to the patient at one time or in a series of administrations to achieve the optimum therapeutic effect.

Depending on the factors mentioned above, the usual dosage ranges may range from about 0.1. mu.g/kg up to about 30mg/kg or more. In particular embodiments, the dosage may range from 1.0 μ g/kg up to about 20mg/kg, optionally from 10 μ g/kg up to about 10mg/kg or from 100 μ g/kg up to about 5 mg/kg.

A therapeutically effective amount of an antibody construct of the invention preferably results in a reduction in the severity of disease symptoms, an increase in the frequency or duration of asymptomatic phases of the disease or the prevention of injury or disability due to the affliction of the disease. For the treatment of EGFRVIII-expressing tumors, a therapeutically effective amount of an antibody construct of the invention, e.g., an anti-EGFRVIII/anti-CD 3 antibody construct, preferably inhibits cell growth or tumor growth by at least about 20%, at least about 40%, at least about 50%, at least about 60%, at least about 70%, at least about 80%, or at least about 90% (relative to untreated patients). The ability of a compound to inhibit tumor growth can be assessed in an animal model that predicts efficacy in human tumors.

The pharmaceutical compositions can be administered as a sole therapeutic agent or in combination with other therapeutic agents such as anti-cancer therapies (e.g., other proteinaceous and non-proteinaceous drugs) as desired. These medicaments may be administered simultaneously with the composition comprising the antibody construct of the invention as defined herein or separately before or after the administration of the antibody construct at defined time intervals and dosages.

As used herein, the term "effective and non-toxic dose" refers to a tolerable dose of the antibody construct of the invention that is sufficiently high to cause pathological cell depletion, tumor elimination, tumor shrinkage or stabilization of the disease, without or substantially without major toxic effects. Such effective and non-toxic doses can be determined, for example, by dose escalation studies described in the art and should be lower than the dose that induces severe adverse side effects (dose-limiting toxicity, DLT).

As used herein, the term "toxicity" refers to the toxic effects of a drug manifested as an adverse event or serious adverse event. These side-events may refer to a lack of global tolerance to the drug and/or a lack of local tolerance after administration. Toxicity may also include teratogenicity or carcinogenesis caused by drugs.

As used herein, the terms "safety", "in vivo safety" or "tolerability" define that administration of a drug does not induce serious adverse events directly after administration (local tolerance) and for an extended period of time after administration. "safety", "in vivo safety" or "tolerability" can be assessed periodically, e.g., during treatment and follow-up. Measurements include clinical assessments, such as the performance of organs, and screening for laboratory abnormalities. Clinical assessments can be performed and deviations from normal outcomes recorded/encoded according to NCI-CTC and/or MedDRA standards. The performance of the organ may include criteria such as allergy/immunology, blood/bone marrow, arrhythmia, coagulation, etc., as set forth in, for example, adverse event terminology standard version 3.0 (CTCAE). Laboratory parameters that may be tested include, for example, hematology, clinical chemistry, coagulation function profiles, and urinalysis, as well as examination of other body fluids such as serum, plasma, lymph or spinal fluid, cerebrospinal fluid, and the like. Thus, the price safety can be evaluated, for example, by physical examination, imaging techniques (i.e. ultrasound, x-ray, CT scan, Magnetic Resonance Imaging (MRI)), other measurements and technical devices (i.e. electrocardiogram), vital signs, by measuring laboratory parameters and recording adverse events. For example, non-chimpanzee primates in the use and method according to the invention can be examined for adverse events by histopathological and/or histochemical methods.

The above terms are also described in, for example, the preliminary safety evaluation of biotechnology-derived pharmaceuticals S6; ICH Harmonied Tripartite Guideline; ICH Steering Committee conference (1997, 16.7.1997).

In further embodiments, the invention provides a host kit comprising an antibody construct of the invention, an antibody construct produced according to a method of the invention, a polynucleotide of the invention, a vector of the invention and/or an invention.

In the context of the present invention, the term "kit" means two or more components packaged together in a container, receptacle or other container, one of which corresponds to an antibody construct, pharmaceutical composition, vector or host cell of the invention. Thus, a kit may be described as a set of products and/or appliances sufficient to achieve a particular goal, which may be sold as a single unit.

The kit may comprise one or more containers (such as vials, ampoules, containers, syringes, bottles, bags) of any suitable shape, size and material (preferably waterproof, e.g. plastic or glass) comprising the antibody construct or pharmaceutical composition of the invention (see above) in a suitable administration dose. The kit may further comprise instructions for use (e.g. in the form of a brochure or instructions), means for administering the antibody construct of the invention (such as a syringe, pump, infuser, etc.), means for reconstituting the antibody construct of the invention and/or means for diluting the antibody construct of the invention.

The invention also provides kits for single dose administration of the units. The kit of the invention may also contain a first receptacle comprising a dried/lyophilized antibody construct and a second receptacle comprising an aqueous formulation. In certain embodiments of the invention, kits containing single-chamber and multi-chamber pre-filled syringes (e.g., liquid syringes and lyosyringes) are provided.

The drawings show that:

FIG. 1:

schematic representation of the tumor specific EGFR mutant EGFRvIII with an N-terminal deletion of 267 amino acids found in glioblastoma.

FIG. 2

The target binder EvIII-2 and its derivatives were compared to the sequence of the covalently linked V-region EvIII-1.

FIG. 3

EvIII-1 and EvIII-2 purification according to standard research grade production

FIG. 4

EvIII-1 and EvIII-2 monomers: reduced SDS-PAGE.

FIG. 5

Cross-reactivity of the EvIII-1 and EvIII-2 bispecific antibody constructs as shown by flow cytometry: binding to human and cynomolgus EGFRvIII and CD3

FIG. 6

Binding of EvIII-1 and EvIII-2 bispecific antibody constructs to EGFRvIII transfected cells and human glioblastoma cell line U87

FIG. 7

Cytotoxic activity of stimulated human CD 8T cells against CHO cells transfected with human EGFRvIII. 18 hours⁵¹And (4) chromium release measurement. Effector cells: stimulated enriched human CD 8T cells. Target cell: CHO cells transfected with EGFRvIII. Effector to target cell (E: T) ratio: 10: 1

FIG. 8

Cytotoxic activity of stimulated human CD 8T cells against human glioblastoma cell line U87: (a)18 hours⁵¹And (4) chromium release measurement. Effector cells: stimulated enriched human CD 8T cells. Target cell: vIII highly positive U87 cells. Effector to target cell (E: T) ratio: 10: 1; (b) FACS-based assay for 18 hours. Effector cells: stimulated enriched human CD 8T cells. Target cell: hu EGFRvIII highly positive U87 glioma cells. Effector to target cell (E: T) ratio: 10: 1.

FIG. 9

Stimulated human CD 8T cells were directed against a human tumor cell line expressing the native EGFRvIII antigen: cytotoxic activity of glioblastoma cell line DK-MG: 18 hours⁵¹And (4) chromium release measurement. Effector cells: stimulated enriched human CD 8T cells. Target cell: DK-MG cells. Effector to target cell (E: T) ratio: 10: 1.

FIG. 10 shows a schematic view of a

Cytotoxic activity of cynomolgus T cell line against CHO cells transfected with cynomolgus EGFRvIII: FACS-based cytotoxicity assay 48 hours. Effector cells: macaque CD3+ LnPx 4119. Target cell: CHO cells transfected with cynomolgus EGFRvIII. Effector to target cell (E: T) ratio: 10: 1.

FIG. 11

Stability of bispecific antibody constructs after 24 hours incubation in human plasma: 18 hours based on⁵¹And (5) measuring Cr. Effector cells: stimulated enriched human CD 8T cells. Target cell: CHO cells transfected with hu EGFRvIII. Effector to target cell (E: T) ratio: 10: 1. Antibody constructs are shown

FIG. 12

Protein homogeneity of EGFRvIII antibody constructs analyzed by high resolution cation exchange chromatography CIEX.

FIG. 13

Surface hydrophobicity of bispecific antibody constructs tested in Hydrophobic Interaction Chromatography (HIC) in flow-through mode.

FIG. 14

Conversion of monomer to dimer after incubation at 250. mu.g/ml and 37 ℃ for 7 days. Analysis was performed by HP-SEC.

FIG. 15 shows a schematic view of a

Conversion of monomer to dimer after three freeze-thaw cycles at 250. mu.g/ml and 37 ℃. Analysis was performed by HP-SEC.

FIG. 16:

FACS binding analysis of the EvIII-1xCD3-scFc construct to CHO cells transfected with human EGFR and the human CD3+ T cell line HPBalL. The red line indicates cells incubated with 2. mu.g/ml purified monomeric protein, which were subsequently incubated with mouse anti-I2C antibody and PE-labeled goat anti-mouse IgG detection antibody. Black bar graph line reflects negative control: cells incubated with anti-12C antibody alone and PE-labeled detection antibody.

FIG. 17:

cytotoxic activity induced by the EvIII-1xCD3-scFc construct redirected to unstimulated human PBMC depleted of CD56 as effector cells and CHO cells transfected with human EGFR as target cells. (example 1.2)

Example (b):

the following examples illustrate the invention. These examples should not be construed as limiting the scope of the invention. The invention is limited only by the claims.

Example 1

Cytotoxic Activity

The efficacy of the egfriiixcd 3 bispecific antibody constructs of the invention to redirect effector T cells against EGFRVIII-expressing target cells was analyzed in five in vitro cytotoxicity assays:

● at 18 hours⁵¹Measurement of the EGFRUIXCD 3 bispecific antibody construct in a Cr Release assay (effective target ratio 10: 1) will stimulate human CD8+ EffectEfficacy of T cells against redirection against CHO cells transfected with human EGFRVIII. FIG. 7

● at 18 hours⁵¹The efficacy of the egfriiixcd 3 bispecific antibody construct to redirect stimulated human CD8+ effector T cells against the EGFRVIII positive human glioblastoma cell line U87 was measured in a Cr release assay (effective target ratio 10: 1). FIG. 8

● measurement of egfriiixcd 3 bispecific antibody constructs human PBMC that would not be stimulated in the absence and presence of soluble EGFRVIII in a 48 hour FACS-based cytotoxicity assay (CD 14)^-/CD56^-) The efficacy of T cells in (1) against human EGFRVIIII transfected CHO cells (10: 1 effective target ratio). FIG. 6 and Table 6

● measurement of EGFRUIdCD 3 bispecific antibody constructs unstimulated human PBMCs (CD 14) in a 48 hour FACS-based cytotoxicity assay^-/CD56^-) The efficacy of T cells in (a) against the EGFRVIII positive human glioblastoma cell line U87 redirection. FIG. 8

● to confirm that the cross-reactive egfriiixcd 3 bispecific antibody construct was able to redirect cynomolgus T cells against cynomolgus EGFRVIII transfected CHO cells, FACS-based cytotoxicity assays were performed for 48 hours using the cynomolgus T cell line LnPx4119 as effector T cells (effective to target ratio 10: 1). FIG. 10

Example 1.1

Chromium release assay using stimulated human T cells

Enriched CD8 was obtained as follows⁺A stimulated T cell of T cells. A cover plate (diameter 145mm, Greiner bio-one GmbH, Kremsm Hunster) was coated with a commercial anti-CD 3 specific antibody (OKT3, Orthoclone) at a final concentration of 1. mu.g/ml for 1 hour at 37 ℃. Unbound protein was removed by one PBS wash step. Will have stabilized glutamine/10% FCS/IL-220U/ml at 120 ml: (

Chiron) 3-5X 10 in RPMI1640⁷Personal PBMCs were added to pre-coated petri dishes and stimulated for 2 days. In thatOn the third day, cells were collected and washed once with RPMI 1640. IL-2 was added to a final concentration of 20U/ml and the cells were cultured for one more day in the same cell culture medium as described above. Consumption of CD4 by Using Dynal-Beads according to the manufacturer's protocol⁺T cells and CD56⁺Enrichment of NK cells for CD8⁺Cytotoxic T Lymphocytes (CTL).

CHO target cells transfected with Cyno EGFRVIII or human EGFRVIII were washed twice with PBS and 11.1MBq in a final volume of 100. mu.l RPMI with 50% FCS at 37 ℃⁵¹Cr mark 60 minutes. Subsequently, the labeled target cells were washed 3 times with 5ml RPMI and then used for cytotoxicity assays. In 96-well plates, measurements were performed in a total volume of 200. mu.l of supplemented RPMI at a 10: 1 ratio of E: T. Purified bispecific antibody constructs and their triplicate dilutions were used at initial concentrations of 0.01-1 μ g/ml. The incubation time of the assay was 18 hours. Cytotoxicity was determined as the relative value of chromium released in the supernatant relative to the difference between maximal lysis (addition of Triton-X) and spontaneous lysis (non-responder cells). All measurements were performed in quadruplicate. In a Wizard 3 "gamma counter (Perkin Elmer Life Sciences GmbH,

germany) was performed. Results analysis was performed using Prism 5 for Windows (version 5.0, GraphPad Software inc., San Diego, California, USA). EC50 values calculated from sigmoidal dose-response curves by analytical procedures were used to compare cytotoxic activity.

Example 1.2

Efficacy of redirecting stimulated human effector T cells against human EGFRVIIII transfected CHO cells

At 51-chromium (⁵¹Cr) release cytotoxicity assay the cytotoxic activity of the egfriiixcd 3 bispecific antibody construct according to the invention was analyzed using human EGFRVIII transfected CHO cells as target cells and stimulated human CD8+ T cells as effector cells. The experiment was performed as described in example 1.1.

The EGFRVIII xcd3 bispecific antibody construct showed very potent cytotoxic activity against human EGFRVIII-transfected CHO cells, ranging from picomolar concentrations at position 1.

Example 1.3

Efficacy of redirecting stimulated human effector T cells against the EGFRVIII positive human cell line human glioblastoma cell line DK-MG

In that⁵¹-chromium (C)⁵¹Cr) cytotoxic activity of the egfriiixcd 3 bispecific antibody construct was analyzed using the EGFRVIII positive human glioblastoma cell line DK-MG as the target cell source and stimulated human CD8+ T cells as effector cells in a cytotoxicity release assay. The assay was performed as described in example 1.1.

Consistent with the results of the 51-chromium release assay using stimulated enriched human CD8+ T lymphocytes as effector cells and human EGFRVIII transfected CHO cells as target cells, the EGFRVIII xcd3 bispecific antibody constructs of the invention also have potent cytotoxic activity against naturally expressing target cells; see fig. 9.

Example 1.4

FACS-based cytotoxicity assays with unstimulated human PBMC

Isolation of Effector cells

Human Peripheral Blood Mononuclear Cells (PBMCs) were prepared from enriched lymphocyte preparations (buffy coats) (by-products of blood banks from which blood for transfusion was collected) by Ficoll density gradient centrifugation. Buffy coat is provided from local blood banks and PBMCs are prepared on the day of blood collection. After Ficoll density centrifugation and extensive washing with Duchen PBS (Gibco), the remaining erythrocytes were passed through a column of erythrocyte lysis buffer (155mM NH)₄Cl、10mM KHCO₃100 μ M EDTA) were removed from the PBMCs by incubation together. After centrifugation of PBMCs at 100xg, platelets were removed via supernatant. The remaining lymphocytes mainly include B and T lymphocytes, NK cells and monocytes. PBMCs were maintained in RPMI Medium (Gibco) containing 10% FCS (Gibco) at 37 ℃/5% CO₂And (5) the following.

CD14⁺And CD56⁺Depletion of cells

In order to consume CD14⁺Cells were used, human CD14 MicroBeads (Milteny Biotec, MACS, # 130-. PBMC were counted and centrifuged at 300Xg for 10 min at room temperature. The supernatant was discarded and the cell pellet was resuspended in MACS separation buffer [ 80. mu.L/10 ]⁷(ii) individual cells; PBS (Invitrogen, #20012-]In (1). CD14 MicroBeads and CD56 MicroBeads (20. mu.L/10) were added⁷Individual cells) and incubated at 4-8 ℃ for 15 minutes. Isolation buffer (1-2 mL/10) with MACS⁷Individual cells) were washed. After centrifugation (see above), the supernatant was discarded and the cells were resuspended in MACS separation buffer (500. mu.L/10)⁸Individual cells). CD14/CD56 negative cells were then isolated using an LS column (Miltenyi Biotec, # 130-. PBMCs without CD14+/CD56+ cells were cultured in RPMI complete medium (i.e., RPMI 1640; Biochrom AG, # FG1215) supplemented with 10% FBS (Biochrom AG, # S0115), 1 × non-essential amino acids (Biochrom AG, # K0293), 10mM Hepes buffer (Biochrom AG, # L1613), 1mM sodium pyruvate (Biochrom AG, # L0473), and 100U/mL penicillin/streptomycin (Biochrom AG, # A2213) in an incubator at 37 ℃ until needed.

Target cell labeling

For analysis of cell lysis in flow cytometry assays, the fluorescent membrane dye DiOC was used₁₈(DiO) (Molecular Probes, # V22886) labeled CHO cells transfected with human EGFRVIII or cynomolgus EGFRVIII as target cells and distinguished from effector cells. Briefly, cells were harvested, washed once with PBS, and washed in a medium containing 2% (v/v) FBS and the membrane dye DiO (5. mu.L/10)⁶Individual cells) was adjusted to 10 in PBS⁶Individual cells/mL. After incubation at 37 ℃ for 3 min, cells were washed twice in complete RPMI medium and cell number was adjusted to 1.25X 10⁵Individual cells/mL. Cell viability was determined using a 0.5% (v/v) isotonic eosin g solution (Roth, # 45380).

Flow cytometry-based analysis

This assay was designed to quantify the lysis of cyno EGFRVIII or human EGFRVIII transfected CHO cells in the presence of serial dilutions of EGFRVIII bispecific antibody constructs. Equal volumes of DiO labeled target and effector cells (i.e., without CD 14)⁺PBMC of cells) to produce an E: T cell ratio of 10: 1. 160 μ l of this suspension was transferred to each well of a 96-well plate. 40 μ L serial dilutions of EGFRUIdCD 3 bispecific antibody construct and negative control bispecific antibody construct (CD 3 based bispecific antibody construct recognizing an unrelated target antigen) or RPMI complete medium (as an additional negative control) were added. Cytotoxicity response mediated by bispecific antibody in 7% CO₂For 48 hours in a humidified incubator. The cells were then transferred to a new 96-well plate and the loss of target cell membrane integrity was monitored by the addition of Propidium Iodide (PI) at a final concentration of 1. mu.g/mL. PI is a membrane-impermeable dye that is normally rejected by living cells, whereas dead cells absorb it and become identifiable by fluorescence emission.

Samples were measured by flow cytometry on a facscan II instrument (from Becton Dickinson) and analyzed by FACSDiva software (from Becton Dickinson). Target cells were identified as DiO positive cells. PI negative target cells were classified as live target cells. Percent cytotoxicity was calculated according to the formula:

n is the number of events

The percent cytotoxicity versus the concentration of the corresponding bispecific antibody construct was plotted using GraphPad Prism 5 Software (GraphPad Software, San Diego). Dose response curves were analyzed using a four-parameter logistic regression model for evaluating sigmoidal dose response curves with fixed hill slopes and EC50 values were calculated.

Example 1.5

Efficacy of redirection of unstimulated human PBMC against human EGFRVIII transfected CHO cells in the absence and presence of soluble EGFRVIIII

The cytotoxic activity of the egfrgixcd 3 bispecific antibody construct was analyzed in FACS-based cytotoxicity assays using human EGFRVIII transfected CHO cells as target cells and unstimulated human PBMC cells as effector cells. The assay was performed as described above in example 1.4.

Unexpectedly, EC50 values were generally higher in cytotoxicity assays using unstimulated PBMC as effector cells compared to cytotoxicity assays using stimulated human CD8+ T cells (see example 1.2).

Example 1.6

Efficacy of redirection of unstimulated human PBMCs against EGFRVIII-positive human glioblastoma cell line U87 or DK-MG cells

In addition, the cytotoxic activity of the EGFRVIII cd3 bispecific antibody construct was analyzed in FACS-based cytotoxicity assays using EGFRVIII positive human glioblastoma cell line U87 or DK-MG as the target cell source and unstimulated human PBMC cells as effector cells. The assay was performed as described above in example 1.4. The results are shown in fig. 8 and 9.

Example 1.7

Effect of redirecting macaque T cells against macaque EGFRVIIII expression CHO cells

The cytotoxic activity of the EGFRVIII cd3 bispecific antibody construct was analyzed in FACS-based cytotoxicity assays using cynomolgus (cyno) EGFRVIII transfected CHO cells as target cells and the cynomolgus T cell line 4119LnPx (Knappe et al, Blood 95: 3256-61(2000)) as effector cell source. Target cell labeling and basing on macaque EGFRVIIII transfected CHO cells as described above⁵¹Analysis of cytotoxic Activity of Cr Release.

The results are shown in FIG. 10. The EGFRVIII xcd3 bispecific antibody construct of the present invention induced macaque T cells from cell line 4119LnPx to effectively kill macaque EGFRVIII transfected CHO cells.

Practice ofExample 1.8

Potency differences between monomeric and dimeric isotypes of bispecific antibody constructs

To determine the difference in cytotoxic activity between the monomeric and dimeric isotypes of a single egfrviii cd3 bispecific antibody construct (referred to as potency differences), an 18 hour 51-chromium release cytotoxicity assay was performed using purified bispecific antibody construct monomers and dimers as described above (example 1.1). The effector cells are stimulated enriched human CD8+ T cells. The target cells were CHO cells transfected with hu EGFRVIII. The ratio of effector cells to target cells (E: T) was 10: 1. The difference in potency was calculated as the ratio between EC50 values.

Example 2

Stability after 24 hours incubation in human plasma

The purified bispecific antibody construct was incubated in a human plasma bank at a ratio of 1: 5 at a final concentration of 2-20. mu.g/ml at 37 ℃ for 96 hours. After incubation with plasma, the antibody constructs were compared in a 51-chromium release assay (as described in example 1.1) using stimulated enriched human CD8+ T cells and human EGFRVIIII transfected CHO cells at an initial concentration of 0.01-0.1 μ g/ml and with an effector to target cell (E: T) ratio of 10: 1. As a control, an unincubated freshly thawed bispecific antibody construct was included.

The results are shown in FIG. 11; the antibody construct EvIII-2 has a plasma stability (EC) of about 2₅₀plasma/EC₅₀Control). Surprisingly, the bispecific antibody constructs of the invention did show little conversion.

Example 3

Analysis of protein homogeneity by high resolution cation exchange chromatography

The protein homogeneity of the antibody constructs of the invention was analyzed by high resolution cation exchange Chromatography (CIEX).

50. mu.g of antibody construct monomer was diluted with 50ml of binding buffer A (20mM sodium dihydrogen phosphate, 30mM NaCl, 0.01% sodium caprylate; pH 5.5),and 40ml of this solution was applied to

Micro FPLC device (GE Healthcare, Germany) in a 1ml BioPro SP-F column (YMC, Germany). After sample binding, a washing step is performed with additional binding buffer. For protein elution, 10 column volumes of buffer B (20mM sodium dihydrogen phosphate, 1000mM NaCl, 0.01% sodium caprylate; pH 5.5) were applied up to a linear increasing salt gradient of 50% buffer B. The entire run was monitored at absorbances of 280, 254 and 210 nm. The analysis is carried out by recording in

The Unicorn software runs the peak integration completion of the 280nm signal in the evaluation table.

The results are shown in FIG. 12. Almost all antibody constructs tested had a very favorable homogeneity (area under the curve of the main peak (═ AUC)) of > 95%.

Example 4

Surface hydrophobicity as measured by HIC Butyl

The surface hydrophobicity of the bispecific antibody constructs of the invention was tested in a flow-through mode in Hydrophobic Interaction Chromatography (HIC).

Mu.g of antibody construct monomer was diluted to a final volume of 500. mu.l with the general formulation buffer (10mM citric acid, 75mM lysine HCl, 4% trehalose; pH 7.0) and applied to

A Purifier FPLC system (GE Healthcare, Germany) was connected to a 1ml butyl Sepharose FF column (GE Healthcare, Germany). The entire run was monitored at absorbances of 280, 254 and 210 nm. The analysis is carried out by recording in

The Unicorn software runs the peak integration completion of the 280nm signal in the evaluation table. Evaluation of elution by comparing the area and speed of the rise and fall of the protein signalBehavior, indicating the strength of the interaction of the BiTE albumin fusion with the matrix.

The antibody constructs had good elution behavior, mostly fast and intact; see fig. 13.

Example 5

Monomer to dimer conversion after (i) three freeze-thaw cycles and (ii) incubation for 7 days at 250 μ g/ml

The bispecific EGFRVIIIxCD3 antibody monomer constructs were subjected to different stress conditions, followed by high-efficiency SEC to determine the percentage of the original monomer antibody construct that had been converted to a dimeric antibody construct.

(i) Mu.g of monomeric antibody construct was adjusted to a concentration of 250. mu.g/ml with the general formulation buffer, then frozen at-80 ℃ for 30 minutes, followed by thawing at room temperature for 30 minutes. After three freeze-thaw cycles, the dimer content was determined by HP-SEC.

(ii) Mu.g of monomeric antibody construct was adjusted to a concentration of 250. mu.g/ml with the general formulation buffer and then incubated at 37 ℃ for 7 days. Dimer content was determined by HP-SEC.

High resolution SEC columns TSK gel G3000 SWXL (Tosoh, Tokyo-Japan) were connected to an A905 autosampler equipped

Purifier 10 FPLC (GE Lifesciences). Column equilibration and running buffer consisted of 100mM KH2PO4-200mM Na2SO4 adjusted to pH 6.6. Antibody solution (25. mu.g protein) was applied to the equilibrated column and eluted at a flow rate of 0.75ml/min at a maximum pressure of 7 MPa. The entire run was monitored at absorbances of 280, 254 and 210 nm. The analysis is carried out by recording in

The Unicorn software runs the peak integration completion of the 210nm signal in the evaluation table. Dimer content was calculated by dividing the area of the dimer peak by the total area of the monomer peak plus dimer peak.

The results are shown in fig. 14 and 15 below. The EVIII-1xCD3 bispecific antibody construct of the invention exhibited a dimer percentage of 0.59% after three freeze-thaw cycles (fig. 15) and 0.26% after 7 days of incubation at 37 ℃, whereas the EVIII-1xCD3 bispecific antibody construct showed a higher value of 1.56% after 7 days of incubation and a higher value of 2.53% after three freeze-thaw cycles.

Example 6

Thermal stability

The antibody aggregation temperature was determined as follows: mu.l of a 250. mu.g/ml solution of the antibody construct was transferred to a disposable cuvette and placed in a Wyatt dynamic light scattering device DynaPro Nanostat (Wyatt). The sample was heated from 40 ℃ to 70 ℃ at a heating rate of 0.5 ℃/min and radius measurements were continuously taken. The increase in radius indicating melting and aggregation of the protein was used by the software package delivered with the DLS device to calculate the aggregation temperature of the antibody construct.

Table 2: thermostability of bispecific antibody constructs determined by DLS (dynamic light Scattering)

EGFRVIII HALB BiTE	Thermostable DLS T_A[℃]
		EvIII-1	51.6
EvIII-2	51.2

Example 7

Turbidity at an antibody concentration of 2500. mu.g/ml

1ml of the purified antibody construct solution at a concentration of 250. mu.g/ml was concentrated to 2500. mu.g/ml by a spin concentration unit. After 16 hours of storage at 5 ℃, the turbidity of the antibody solution was determined by OD340nm absorbance measurements against the universal formulation buffer.

The results are shown in table 3 below. The EvIII-1 antibody constructs of the invention have a very favourable turbidity of ≦ 0.03, whereas the EvIII-2 antibody constructs exhibit a pronounced turbidity, indicating less favourable properties when formulating this molecule in pharmaceutical compositions.

Table 3: turbidity of antibody construct after overnight concentration to 2.5mg/ml

Claims

1. A bispecific antibody construct comprising (i) a first binding domain that binds to human and cynomolgus EGFRVIII on the surface of a target cell, said first binding domain comprising a VH region and a VL region, and (ii) a second binding domain that binds to human CD3 on the surface of a T cell, said second binding domain comprising a VH region and a VL region, wherein the amino acid sequence of the VH region of said first binding domain is set forth in SEQ ID NO:157 and the amino acid sequence of the VL region of said first binding domain is set forth in SEQ ID NO: 158.

2. The antibody construct according to claim 1, wherein the antibody construct is in a form selected from the group consisting of: (scFv)₂scFv-single domain mAbs, diabodies and oligomers of any of these forms.

3. The antibody construct according to claim 1, wherein the second binding domain binds to human CD3 epsilon and to Callithrix jacchus, Tamarix villosa or Saimiri sciureus CD3 epsilon.

4. The antibody construct according to claim 1, wherein the second binding domain binds to human CD3 epsilon and to Callithrix jacchus, Tamarix villosa or Saimiri sciureus CD3 epsilon and comprises:

(i) a VL region comprising CDR-L1 as shown in SEQ ID NO. 11, CDR-L2 as shown in SEQ ID NO. 12 and CDR-L3 as shown in SEQ ID NO. 13; and a VH region comprising CDR-H1 as shown in SEQ ID NO. 14, CDR-H2 as shown in SEQ ID NO. 15, and CDR-H3 as shown in SEQ ID NO. 16;

(ii) a VL region comprising CDR-L1 as shown in SEQ ID NO:20, CDR-L2 as shown in SEQ ID NO:21 and CDR-L3 as shown in SEQ ID NO: 22; and a VH region comprising CDR-H1 as shown in SEQ ID NO:23, CDR-H2 as shown in SEQ ID NO:24, and CDR-H3 as shown in SEQ ID NO: 25;

(iii) a VL region comprising CDR-L1 as shown in SEQ ID NO. 29, CDR-L2 as shown in SEQ ID NO. 30 and CDR-L3 as shown in SEQ ID NO. 31; and a VH region comprising the CDR-H1 shown in SEQ ID NO. 32, the CDR-H2 shown in SEQ ID NO. 33, and the CDR-H3 shown in SEQ ID NO. 34;

(iv) a VL region comprising CDR-L1 as shown in SEQ ID NO:38, CDR-L2 as shown in SEQ ID NO:39 and CDR-L3 as shown in SEQ ID NO: 40; and a VH region comprising CDR-H1 as shown in SEQ ID NO:41, CDR-H2 as shown in SEQ ID NO:42, and CDR-H3 as shown in SEQ ID NO: 43;

(v) a VL region comprising CDR-L1 as shown in SEQ ID NO:47, CDR-L2 as shown in SEQ ID NO:48 and CDR-L3 as shown in SEQ ID NO: 49; and a VH region comprising CDR-H1 as shown in SEQ ID NO:50, CDR-H2 as shown in SEQ ID NO:51, and CDR-H3 as shown in SEQ ID NO: 52;

(vi) a VL region comprising CDR-L1 as shown in SEQ ID NO:56, CDR-L2 as shown in SEQ ID NO:57, and CDR-L3 as shown in SEQ ID NO: 58; and a VH region comprising CDR-H1 as shown in SEQ ID NO:59, CDR-H2 as shown in SEQ ID NO:60, and CDR-H3 as shown in SEQ ID NO: 61;

(vii) a VL region comprising CDR-L1 as shown in SEQ ID NO:65, CDR-L2 as shown in SEQ ID NO:66 and CDR-L3 as shown in SEQ ID NO: 67; and a VH region comprising CDR-H1 as shown in SEQ ID NO:68, CDR-H2 as shown in SEQ ID NO:69, and CDR-H3 as shown in SEQ ID NO: 70;

(viii) a VL region comprising CDR-L1 as shown in SEQ ID NO:74, CDR-L2 as shown in SEQ ID NO:75 and CDR-L3 as shown in SEQ ID NO: 76; and a VH region comprising CDR-H1 as shown in SEQ ID NO:77, CDR-H2 as shown in SEQ ID NO:78, and CDR-H3 as shown in SEQ ID NO: 79;

(ix) a VL region comprising CDR-L1 as shown in SEQ ID NO:83, CDR-L2 as shown in SEQ ID NO:84 and CDR-L3 as shown in SEQ ID NO: 85; and a VH region comprising CDR-H1 as shown in SEQ ID NO:86, CDR-H2 as shown in SEQ ID NO:87, and CDR-H3 as shown in SEQ ID NO: 88;

(x) A VL region comprising CDR-L1 as shown in SEQ ID NO:92, CDR-L2 as shown in SEQ ID NO:93 and CDR-L3 as shown in SEQ ID NO: 94; and a VH region comprising CDR-H1 as shown in SEQ ID NO:95, CDR-H2 as shown in SEQ ID NO:96, and CDR-H3 as shown in SEQ ID NO: 97; or

(xi) A VH region and a VL region, wherein the amino acid sequence of the VH region is shown as SEQ ID NO. 101, and the amino acid sequence of the VL region is shown as SEQ ID NO. 102.

5. The antibody construct of claim 1, comprising:

(a) a polypeptide comprising, in order from the N-terminus:

(i) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO. 159;

(ii) a peptide linker, wherein the amino acid sequence of the peptide linker is selected from the amino acid sequences as set forth in any one of SEQ ID NOs 1-9; and

(iii) a polypeptide, wherein the amino acid sequence of the polypeptide is selected from the group consisting of the amino acid sequences shown as SEQ ID NO 19, SEQ ID NO 28, SEQ ID NO 37, SEQ ID NO 46, SEQ ID NO 55, SEQ ID NO 64, SEQ ID NO 73, SEQ ID NO 82, SEQ ID NO 91, SEQ ID NO 100 and SEQ ID NO 103; and

(iv) optionally, a His-tag, wherein the amino acid sequence of the His-tag is set forth in SEQ ID NO 10;

(b) a polypeptide comprising, in order from the N-terminus:

(ii) a peptide linker, wherein the amino acid sequence of the peptide linker is selected from the amino acid sequences as set forth in any one of SEQ ID NOs 1-9;

(iii) a polypeptide, wherein the amino acid sequence of the polypeptide is selected from the group consisting of the amino acid sequences shown as SEQ ID NO 19, SEQ ID NO 28, SEQ ID NO 37, SEQ ID NO 46, SEQ ID NO 55, SEQ ID NO 64, SEQ ID NO 73, SEQ ID NO 82, SEQ ID NO 91, SEQ ID NO 100 and SEQ ID NO 103;

(iv) optionally, a peptide linker, wherein the amino acid sequence of the peptide linker is selected from the amino acid sequences as set forth in any one of SEQ ID NOs 1-9;

(v) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO: 134; and

(vi) optionally, a His-tag, wherein the amino acid sequence of the His-tag is set forth in SEQ ID NO 10;

(c) a polypeptide comprising, in order from the N-terminus:

(i) polypeptide, wherein the amino acid sequence of said polypeptide is as the amino acid sequence QRFVTGHFGGLX₁PANG (SEQ ID NO:135) wherein X₁Is Y or H;

(ii) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO. 159;

(iii) a peptide linker, wherein the amino acid sequence of the peptide linker is selected from the amino acid sequences as set forth in any one of SEQ ID NOs 1-9;

(iv) a polypeptide, wherein the amino acid sequence of the polypeptide is selected from the group consisting of the amino acid sequences shown as SEQ ID NO 19, SEQ ID NO 28, SEQ ID NO 37, SEQ ID NO 46, SEQ ID NO 55, SEQ ID NO 64, SEQ ID NO 73, SEQ ID NO 82, SEQ ID NO 91, SEQ ID NO 100 and SEQ ID NO 103;

(v) a polypeptide, wherein the amino acid sequence of the polypeptide is represented by amino acid sequence QRFVTGHFGGLHPANG (SEQ ID NO:137) or QRFCTGHFGGLHPCNG (SEQ ID NO: 139); and

(vi) optionally, a His-tag, the amino acid sequence of which is set forth in SEQ ID NO 10;

(d) a first polypeptide comprising, in order from the N-terminus:

(i) a polypeptide, wherein the amino acid sequence of the polypeptide is selected from the group consisting of the amino acid sequences shown as SEQ ID NO 17, SEQ ID NO 26, SEQ ID NO 35, SEQ ID NO 44, SEQ ID NO 53, SEQ ID NO 62, SEQ ID NO 71, SEQ ID NO 80, SEQ ID NO 89, SEQ ID NO 98 and SEQ ID NO 101;

(ii) a peptide linker, wherein the amino acid sequence of the peptide linker is shown as SEQ ID NO. 8;

(iii) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO: 158; and

(iv) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO: 140;

and a second polypeptide comprising, in order from the N-terminus:

(i) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO: 157;

(iii) a polypeptide, wherein the amino acid sequence of the polypeptide consists of an amino acid sequence selected from the group consisting of SEQ ID NO 18, SEQ ID NO 27, SEQ ID NO 36, SEQ ID NO 45, SEQ ID NO 54, SEQ ID NO 63, SEQ ID NO 72, SEQ ID NO 81, SEQ ID NO 90, SEQ ID NO 99 and SEQ ID NO 102 and a serine residue at the C-terminus; and

(iv) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO. 141;

(e) a first polypeptide comprising, in order from the N-terminus:

(i) a polypeptide, wherein the amino acid sequence of the polypeptide is selected from the group consisting of SEQ ID NO 17, SEQ ID NO 26, SEQ ID NO 35, SEQ ID NO 44, SEQ ID NO 53, SEQ ID NO 62, SEQ ID NO 71, SEQ ID NO 80, SEQ ID NO 89, SEQ ID NO 98 and SEQ ID NO 101;

(ii) a peptide linker, wherein the amino acid sequence of the peptide linker is shown in SEQ ID NO 8;

(iv) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO: 142;

and a second polypeptide comprising, in order from the N-terminus:

(iv) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO. 143;

(f) a first polypeptide comprising, in order from the N-terminus:

(iv) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO: 144;

and a second polypeptide, the amino acid sequence of the second polypeptide is shown as SEQ ID NO. 145;

(g) a first polypeptide comprising, in order from the N-terminus:

(i) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO. 159; and

(ii) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO: 146;

and a second polypeptide comprising, in order from the N-terminus:

(i) a polypeptide, wherein the amino acid sequence of the polypeptide is selected from the group consisting of the amino acid sequences shown as SEQ ID NO 19, SEQ ID NO 28, SEQ ID NO 37, SEQ ID NO 46, SEQ ID NO 55, SEQ ID NO 64, SEQ ID NO 73, SEQ ID NO 82, SEQ ID NO 91, SEQ ID NO 100 and SEQ ID NO 103; and

(ii) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO: 147;

(h) a first polypeptide comprising, in order from the N-terminus:

(ii) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO. 148;

and a second polypeptide comprising, in order from the N-terminus:

(ii) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO: 149;

(i) a polypeptide comprising, in order from the N-terminus:

(iv) a polypeptide, wherein the amino acid sequence of the polypeptide is shown as SEQ ID NO: 150; or

(j) A polypeptide comprising, in order from the N-terminus:

(iv) a peptide linker, wherein the amino acid sequence of the peptide linker is selected from the group consisting of the amino acid sequences shown as SEQ ID NOs 1, 2, 4,5, 6,8, and 9; and

(v) a third domain having an amino acid sequence selected from the amino acid sequences as set forth in any one of SEQ ID NO 181-188.

6. The antibody construct according to claim 1, comprising or consisting of the polypeptide as set forth in SEQ ID NO 160.

7. A polypeptide, comprising:

(a) a first binding domain that binds to human and cynomolgus epidermal growth factor receptor viii (egfrviii) and comprises: a heavy chain variable region (VH) and a light chain variable region (VL), wherein the amino acid sequence of the VH is shown as SEQ ID NO:157, and the amino acid sequence of the VL is shown as SEQ ID NO: 158; and

(b) a second binding domain that binds to human CD3 and comprises: a VL region comprising CDR-L1 as shown in SEQ ID NO:92, CDR-L2 as shown in SEQ ID NO:93 and CDR-L3 as shown in SEQ ID NO: 94; the VH region comprises CDR-H1 shown as SEQ ID NO:95, CDR-H2 shown as SEQ ID NO:96 and CDR-H3 shown as SEQ ID NO: 97.

8. The polypeptide of claim 7, comprising: (a) a first binding domain that binds to human and cynomolgus EGFRVIIII and comprises: VH and VL, wherein the amino acid sequence of VH of the first binding domain is shown as SEQ ID NO:157, and the amino acid sequence of VL of the first binding domain is shown as SEQ ID NO: 158; and (b) a second binding domain that binds to human CD3 and comprises: VH and VL, wherein the amino acid sequence of VH of the second binding domain is shown as SEQ ID NO:98, and the amino acid sequence of VL of the second binding domain is shown as SEQ ID NO: 99.

9. The polypeptide of claim 7, comprising: (a) a first binding domain having the amino acid sequence set forth in SEQ ID NO: 159; and (b) a second binding domain having the amino acid sequence set forth in SEQ ID NO: 100.

10. The polypeptide of claim 7, further comprising a His-tag having an amino acid sequence as set forth in SEQ ID NO 10.

11. The polypeptide of claim 7, wherein the amino acid sequence of the polypeptide comprises the amino acid sequence set forth in SEQ ID NO 160.

12. The polypeptide of claim 11, further comprising a His-tag having the amino acid sequence set forth in SEQ ID No. 10.

13. The polypeptide of claim 11, wherein the amino acid sequence of the polypeptide comprises, in order from the N-terminus: an amino acid sequence shown as SEQ ID NO. 160, and a His tag amino acid sequence shown as SEQ ID NO. 10.

14. The polypeptide of claim 11, further comprising a single chain Fc (scFc), wherein the amino acid sequence of the scFc is selected from the amino acid sequences as set forth in any one of SEQ ID NO 181-188.

15. The polypeptide of claim 11, wherein the amino acid sequence of the polypeptide is set forth in SEQ ID NO: 189.

16. The polypeptide of claim 11, wherein the amino acid sequence of the polypeptide is set forth in SEQ ID NO 190.

17. A polynucleotide encoding an antibody construct as defined in any one of claims 1 to 6 or a polypeptide as defined in any one of claims 7 to 16.

18. A vector comprising a polynucleotide as defined in claim 17.

19. A host cell transformed or transfected with a vector as defined in claim 18.

20. A method of producing an antibody construct according to any one of claims 1 to 6 or a polypeptide as defined in any one of claims 7 to 16, the method comprising culturing a host cell as defined in claim 19 under conditions which allow expression of an antibody construct as defined in any one of claims 1 to 6 or a polypeptide as defined in any one of claims 7 to 16, and recovering the produced antibody construct or polypeptide from the culture.

21. A pharmaceutical composition comprising an antibody construct according to any one of claims 1 to 6 or a polypeptide according to any one of claims 7-16.

22. Use of an antibody construct according to any one of claims 1 to 6 or a polypeptide according to any one of claims 7-16 for the preparation of a pharmaceutical composition for the treatment or amelioration of a tumor or cancer disease or a metastatic cancer disease.

23. The use of claim 22, wherein the tumor or cancer disease is selected from the group consisting of: glioblastoma, astrocytoma, medulloblastoma, breast cancer, non-small cell lung cancer, ovarian cancer, prostate cancer, central nervous system cancer, or metastatic cancer disease derived from any of the foregoing.

24. The use of claim 23, wherein the tumor or cancer disease is glioblastoma.

25. A kit comprising an antibody construct according to any one of claims 1 to 6 or a polypeptide according to any one of claims 7-16.